Re: [OpenIndiana-discuss] Scrub on single ZFS volume
On Mon, 2023-10-23 at 10:30 +0200, Marcel Telka wrote: > > Does a scrub still make sense when there isn't any parity? > Yes > From the zpool man page: > > zpool scrub [-s | -p] pool... > Begins a scrub or resumes a paused scrub. The scrub > examines all > data in the specified pools to verify that it checksums > correctly. > For replicated (mirror or raidz) devices, ZFS > automatically repairs > any damage discovered during the scrub. The zpool > status command > reports the progress of the scrub and summarizes the > results of the > scrub upon completion. > OK - so basically, even though it can't repair damage, (as there is no parity) at least it can flag up that checksums didn't check. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Scrub on single ZFS volume
I'm having problems getting this question answered via the usual google search, so I thought I'd ask here. Situation... Server has three drives in ZFS RAID... they're not the problem. I have external USB drives that I hook up, in order to take a backup of the data. It is also ZFS. (I actually have smaller drives hooked together in RAID 0 to cope with the large amount of data.) The question is... how do I check the integrity of the data on the single backup volume? Does a scrub still make sense when there isn't any parity? Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
Test run again with wireshark on the client... http://www.msknight.com/files/wireshark.pcapng Server is 192.168.0.2 and the test is a file, open in Bluefish, is altered and then asked to save. Bluefish doesn't give an error, but the operation fails and the temporary save file is left intact. Michelle. On Sun, 2023-10-01 at 16:15 +0300, Toomas Soome via openindiana-discuss wrote: > > > > On 1. Oct 2023, at 15:32, Michelle wrote: > > > > http://www.msknight.com/files/capture > > Wireshark is suggesting this file is damaged or corrupt. size is > 63805 bytes. > > rgds, > toomas > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
It was captured with snoop, so presumably you'd need snoop -i to read it? On Sun, 2023-10-01 at 16:15 +0300, Toomas Soome via openindiana-discuss wrote: > > > > On 1. Oct 2023, at 15:32, Michelle wrote: > > > > http://www.msknight.com/files/capture > > Wireshark is suggesting this file is damaged or corrupt. size is > 63805 bytes. > > rgds, > toomas > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
Hopefully, this contains what's needed... http://www.msknight.com/files/capture Michelle. On Mon, 2023-09-25 at 11:39 +0200, Udo Grabowski (IMK) wrote: > On 25/09/2023 11:36, Michelle wrote: > > ... > > > > I'll have to look up snoop and how to use it. ... > > > > snoop -d > > -v for detailed package content. > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
nbmand is set to on I'll have to look up snoop and how to use it. Won't get the time to sit down and learn for a couple of days, but replicating the problem is easy. On Mon, 2023-09-25 at 12:09 +0300, Toomas Soome via openindiana-discuss wrote: > > > > On 25. Sep 2023, at 10:29, Michelle wrote: > > > > Apologies for the delay. Chores last night and just about to start > > work. > > > > Previously I had forced the version to 1.0 and then had to step up > > for > > some reason. I now don't specify a version on the mount and both > > servers connect on 3.1.1 > > > > I think that the current problem was actually causing me to think > > that > > the version number was the problem... until I'd exhausted all > > versions... but given that experience, I doubt that forcing version > > numbers will help here. > > > > > ok, well, this output means, your server is running defaults, and I > was asking about this because you had this message about 0xFE - not > sure if that may or may not have any impact on actual session. It may > just be some sort of informational message to confuse poor users;) > > Now, the actual error is concerning locking, I assume you have > nbmand=on on that share, but it wont hurt to check it. Otherwise we > would need network packet trace from the moment this error is > happening, so we could check what request-response information is > being passed. > > Packet trace could be saved with snoop -O prefix:count:size host > hostname and port 445, but the tricky part there is to set number of > snoop output files + size and stopping the snoop to make sure the > error moment is actually present in recorded file(s). So you want to > prepare the conditions, then start snoop, re-create the error and > stop snoop. > > rgds, > toomas > > > > root@jaguar:~# sharectl get smb > > system_comment= > > max_workers=1024 > > netbios_enable=false > > netbios_scope= > > lmauth_level=4 > > keep_alive=0 > > wins_server_1= > > wins_server_2= > > wins_exclude= > > signing_enabled=true > > signing_required=true > > restrict_anonymous=false > > pdc= > > ads_site= > > ddns_enable=false > > autohome_map=/etc > > ipv6_enable=false > > print_enable=false > > traverse_mounts=true > > map= > > unmap= > > disposition= > > min_protocol= > > max_protocol= > > encrypt=disabled > > encrypt_ciphers= > > bypass_traverse_checking=true > > oplock_enable=true > > short_names=false > > > > > > > > > > > > On Sun, 2023-09-24 at 23:10 +0300, Toomas Soome via openindiana- > > discuss > > wrote: > > > > > > Could you paste: 'sharectl get smb’. The ‘unknown response type > > > 0xfe’ > > > is hinting that your client is expecting 0xFF there, and you > > > probably > > > need to set server to default to older SMB version. > > > > > > rgds, > > > toomas > > > > > > > > > > On 24. Sep 2023, at 22:37, Toomas Soome via openindiana-discuss > > > > wrote: > > > > > > > > > > > > > > On 24. Sep 2023, at 22:34, Michelle > > > > > mailto:miche...@msknight.com> > > > > > <mailto:miche...@msknight.com>> wrote: > > > > > > > > > > I just updated all packages and tested again... > > > > > > > > > > Sep 24 20:31:37 mich-desktop kernel: [311047.112078] CIFS: > > > > > VFS: > > > > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > > > > Sep 24 20:31:37 mich-desktop kernel: [311047.112114] CIFS: > > > > > VFS: > > > > > Push > > > > > locks rc = -22 > > > > > > > > > > I'm at the end. Nowhere else to go. > > > > > > > > > > Michelle. > > > > > > > > on OI, what you get from 'pkg update -nv’ ? > > > > > > > > rgds, > > > > toomas > > > > > > > > > > > > > > On Sun, 2023-09-24 at 20:45 +0300, Toomas Soome via > > > > > openindiana- > > > > > discuss > > > > > wrote: > > > > > > Ou that is sad that hear. Btw, what is the oi build you are > > > > > > using on > > > > > > that server? > > > > > > > > > > > > Sent from my iPhone > > &g
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
Good suggestion. I'll research NFS again, see if I can remember where I got to, and give it another try. On Sun, 2023-09-24 at 22:10 +0200, Goetz T. Fischer wrote: > mount -o ro > > On Sun, 24 Sep 2023 20:41:47 +0100, Michelle wrote: > > I mount a share in read only, and then when I want to write to an > > area > > I unmount it and re-mount it as a different user... and when I'm > > done > > writing to it, then I unmount it and re-mount it as read only > > again... > > to prevent accidental deletions, etc. > > > > I don't think I could do that in NFS. > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
True, but if the problem is not investigated then it will never be solved. On Sun, 2023-09-24 at 22:12 +0200, Goetz T. Fischer wrote: > sure, but the key is WHEN the update happens. the problem might not > exist anymore at that point. > > On Sun, 24 Sep 2023 20:36:21 +0100, Michelle wrote: > > I can only remain on a version for so long. Eventually, updates > > have to > > happen > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
Apologies for the delay. Chores last night and just about to start work. Previously I had forced the version to 1.0 and then had to step up for some reason. I now don't specify a version on the mount and both servers connect on 3.1.1 I think that the current problem was actually causing me to think that the version number was the problem... until I'd exhausted all versions... but given that experience, I doubt that forcing version numbers will help here. root@jaguar:~# sharectl get smb system_comment= max_workers=1024 netbios_enable=false netbios_scope= lmauth_level=4 keep_alive=0 wins_server_1= wins_server_2= wins_exclude= signing_enabled=true signing_required=true restrict_anonymous=false pdc= ads_site= ddns_enable=false autohome_map=/etc ipv6_enable=false print_enable=false traverse_mounts=true map= unmap= disposition= min_protocol= max_protocol= encrypt=disabled encrypt_ciphers= bypass_traverse_checking=true oplock_enable=true short_names=false On Sun, 2023-09-24 at 23:10 +0300, Toomas Soome via openindiana-discuss wrote: > > Could you paste: 'sharectl get smb’. The ‘unknown response type 0xfe’ > is hinting that your client is expecting 0xFF there, and you probably > need to set server to default to older SMB version. > > rgds, > toomas > > > > On 24. Sep 2023, at 22:37, Toomas Soome via openindiana-discuss > > wrote: > > > > > > > > On 24. Sep 2023, at 22:34, Michelle > > > mailto:miche...@msknight.com>> wrote: > > > > > > I just updated all packages and tested again... > > > > > > Sep 24 20:31:37 mich-desktop kernel: [311047.112078] CIFS: VFS: > > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > > Sep 24 20:31:37 mich-desktop kernel: [311047.112114] CIFS: VFS: > > > Push > > > locks rc = -22 > > > > > > I'm at the end. Nowhere else to go. > > > > > > Michelle. > > > > on OI, what you get from 'pkg update -nv’ ? > > > > rgds, > > toomas > > > > > > > > On Sun, 2023-09-24 at 20:45 +0300, Toomas Soome via openindiana- > > > discuss > > > wrote: > > > > Ou that is sad that hear. Btw, what is the oi build you are > > > > using on > > > > that server? > > > > > > > > Sent from my iPhone > > > > > > > > > On 24. Sep 2023, at 17:34, Michelle > > > > > wrote: > > > > > > > > > > I would like to give my thanks to all who have run and > > > > > maintained > > > > > OpenIndiana for the many years it has been available. > > > > > > > > > > I have been using it since big red took the reins of Sun and > > > > > changed > > > > > the license of Open Solaris, as it gave me access to ZFS; a > > > > > product > > > > > that has made a big difference in my personal computing life. > > > > > > > > > > The community has been welcoming and patient with me despite > > > > > my > > > > > novice > > > > > questions. I just wish I had the skills and time to be able > > > > > to > > > > > contribute to the project but my job, my caring duties and > > > > > looking > > > > > after a home have left me with little personal time... > > > > > especially > > > > > in > > > > > this last few years. > > > > > > > > > > Unfortunately this latest issue with CIFS/SMB has been quite > > > > > a > > > > > disruption and with little way to make progress, I’ve > > > > > determined > > > > > that > > > > > as the problems have persisted despite changing the client > > > > > from > > > > > Mint to > > > > > Ubuntu Cinnamon, I now have to change server platform. I’ll > > > > > be > > > > > replacing my OI servers with Debian over the coming weeks. > > > > > > > > > > Thank you all for your patience and help over the years. > > > > > > > > > > Michelle. > > > > > > > > > > > On Tue, 2023-09-19 at 20:54 +0100, Michelle wrote: > > > > > > OK, I've now upgraded the client to Ubuntu Cinnamon 22.04 > > > > > > > > > > > > Accessing the text file on the OI share via BlueFish this > > > > > > time, > > > > > > and > > > > > >
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
root@jaguar:~# pkg update -nv No updates available for this image. On Sun, 2023-09-24 at 22:37 +0300, Toomas Soome via openindiana-discuss wrote: > > > On 24. Sep 2023, at 22:34, Michelle wrote: > > > > I just updated all packages and tested again... > > > > Sep 24 20:31:37 mich-desktop kernel: [311047.112078] CIFS: VFS: > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > Sep 24 20:31:37 mich-desktop kernel: [311047.112114] CIFS: VFS: > > Push > > locks rc = -22 > > > > I'm at the end. Nowhere else to go. > > > > Michelle. > > on OI, what you get from 'pkg update -nv’ ? > > rgds, > toomas > > > > > On Sun, 2023-09-24 at 20:45 +0300, Toomas Soome via openindiana- > > discuss > > wrote: > > > Ou that is sad that hear. Btw, what is the oi build you are using > > > on > > > that server? > > > > > > Sent from my iPhone > > > > > > > On 24. Sep 2023, at 17:34, Michelle > > > > wrote: > > > > > > > > I would like to give my thanks to all who have run and > > > > maintained > > > > OpenIndiana for the many years it has been available. > > > > > > > > I have been using it since big red took the reins of Sun and > > > > changed > > > > the license of Open Solaris, as it gave me access to ZFS; a > > > > product > > > > that has made a big difference in my personal computing life. > > > > > > > > The community has been welcoming and patient with me despite my > > > > novice > > > > questions. I just wish I had the skills and time to be able to > > > > contribute to the project but my job, my caring duties and > > > > looking > > > > after a home have left me with little personal time... > > > > especially > > > > in > > > > this last few years. > > > > > > > > Unfortunately this latest issue with CIFS/SMB has been quite a > > > > disruption and with little way to make progress, I’ve > > > > determined > > > > that > > > > as the problems have persisted despite changing the client from > > > > Mint to > > > > Ubuntu Cinnamon, I now have to change server platform. I’ll be > > > > replacing my OI servers with Debian over the coming weeks. > > > > > > > > Thank you all for your patience and help over the years. > > > > > > > > Michelle. > > > > > > > > > On Tue, 2023-09-19 at 20:54 +0100, Michelle wrote: > > > > > OK, I've now upgraded the client to Ubuntu Cinnamon 22.04 > > > > > > > > > > Accessing the text file on the OI share via BlueFish this > > > > > time, > > > > > and > > > > > making changes and saving, I get (in syslog)... > > > > > > > > > > Sep 19 20:49:47 mich-desktop kernel: [13839.060593] CIFS: > > > > > VFS: > > > > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > > > > Sep 19 20:49:47 mich-desktop kernel: [13839.060714] CIFS: > > > > > VFS: > > > > > Push > > > > > locks rc = -22 > > > > > > > > > > Same behaviour as before when using xed on Mint. If I delete > > > > > and > > > > > then > > > > > save, it works. And if I attempt to rename a file that's in > > > > > use > > > > > (eg. > > > > > a > > > > > video) then I get resource unavailable. > > > > > > > > > > I have no realistic option now other than to conclude that > > > > > it's > > > > > the > > > > > OpenIndiana side of things. > > > > > > > > > > I'd be grateful for guidance on whatever I need to file a bug > > > > > report > > > > > please. > > > > > > > > > > > > > > > > > > > > > On Mon, 2023-08-07 at 22:54 +0100, Michelle wrote: > > > > > > Apparmor is, indeed running, so I created a profile and put > > > > > > it > > > > > > to > > > > > > complain mode... noting that Libre Office is also in > > > > > > complain > > > > > > mode > > > > > > already. > > > > > > > > >
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
I think I recall... there's something I do which can't be done in NFS. I mount a share in read only, and then when I want to write to an area I unmount it and re-mount it as a different user... and when I'm done writing to it, then I unmount it and re-mount it as read only again... to prevent accidental deletions, etc. I don't think I could do that in NFS. On Sun, 2023-09-24 at 21:27 +0200, Goetz T. Fischer wrote: > i didn't quite get the reason why you wanna keep using samba instead > of nfs but that aside, since this > is your home server, why not just keeping the indiana version that > worked? or at least the samba version > that worked? > security usually isn't that big of an issue for this purpose. > > On Sun, 24 Sep 2023 15:34:32 +0100, Michelle wrote: > > I would like to give my thanks to all who have run and maintained > > OpenIndiana for the many years it has been available. > > > > I have been using it since big red took the reins of Sun and > > changed > > the license of Open Solaris, as it gave me access to ZFS; a product > > that has made a big difference in my personal computing life. > > > > The community has been welcoming and patient with me despite my > > novice > > questions. I just wish I had the skills and time to be able to > > contribute to the project but my job, my caring duties and looking > > after a home have left me with little personal time... especially > > in > > this last few years. > > > > Unfortunately this latest issue with CIFS/SMB has been quite a > > disruption and with little way to make progress, I’ve determined > > that > > as the problems have persisted despite changing the client from > > Mint to > > Ubuntu Cinnamon, I now have to change server platform. I’ll be > > replacing my OI servers with Debian over the coming weeks. > > > > Thank you all for your patience and help over the years. > > > > Michelle. > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
By the way, while xed gives an error... if editing a file in Bluefish and saving it to the share, there is the syslog entry but there's no actual warning to the user that the file hasn't saved. It's a case of open a file in bluefish, change it, save it, then open it again and the changes aren't there... that's a scary situation that I can't cope with any more. On Sun, 2023-09-24 at 20:45 +0300, Toomas Soome via openindiana-discuss wrote: > Ou that is sad that hear. Btw, what is the oi build you are using on > that server? > > Sent from my iPhone > > > On 24. Sep 2023, at 17:34, Michelle wrote: > > > > I would like to give my thanks to all who have run and maintained > > OpenIndiana for the many years it has been available. > > > > I have been using it since big red took the reins of Sun and > > changed > > the license of Open Solaris, as it gave me access to ZFS; a product > > that has made a big difference in my personal computing life. > > > > The community has been welcoming and patient with me despite my > > novice > > questions. I just wish I had the skills and time to be able to > > contribute to the project but my job, my caring duties and looking > > after a home have left me with little personal time... especially > > in > > this last few years. > > > > Unfortunately this latest issue with CIFS/SMB has been quite a > > disruption and with little way to make progress, I’ve determined > > that > > as the problems have persisted despite changing the client from > > Mint to > > Ubuntu Cinnamon, I now have to change server platform. I’ll be > > replacing my OI servers with Debian over the coming weeks. > > > > Thank you all for your patience and help over the years. > > > > Michelle. > > > > > On Tue, 2023-09-19 at 20:54 +0100, Michelle wrote: > > > OK, I've now upgraded the client to Ubuntu Cinnamon 22.04 > > > > > > Accessing the text file on the OI share via BlueFish this time, > > > and > > > making changes and saving, I get (in syslog)... > > > > > > Sep 19 20:49:47 mich-desktop kernel: [13839.060593] CIFS: VFS: > > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > > Sep 19 20:49:47 mich-desktop kernel: [13839.060714] CIFS: VFS: > > > Push > > > locks rc = -22 > > > > > > Same behaviour as before when using xed on Mint. If I delete and > > > then > > > save, it works. And if I attempt to rename a file that's in use > > > (eg. > > > a > > > video) then I get resource unavailable. > > > > > > I have no realistic option now other than to conclude that it's > > > the > > > OpenIndiana side of things. > > > > > > I'd be grateful for guidance on whatever I need to file a bug > > > report > > > please. > > > > > > > > > > > > > On Mon, 2023-08-07 at 22:54 +0100, Michelle wrote: > > > > Apparmor is, indeed running, so I created a profile and put it > > > > to > > > > complain mode... noting that Libre Office is also in complain > > > > mode > > > > already. > > > > > > > > However, xed still fails with the same issue. It is, > > > > apparently, > > > > writing a temporary file and presumably failing... the same way > > > > that > > > > copying a file from the local machine to the share is > > > > failing... > > > > > > > > ... copying the file to a temporary .gooutputstream file but > > > > then > > > > somehow failing to delete the file that is about to be > > > > replaced. > > > > > > > > Michelle. > > > > > > > > On Mon, 2023-08-07 at 22:24 +0100, Michelle wrote: > > > > > I'm starting to believe that upgrading Linux has installed > > > > > apparmor > > > > > ... > > > > > and it's likely that apparmor is now getting in the way of > > > > > various > > > > > applications. > > > > > > > > > > I hate apparmor. I mean, I understand why it's there and the > > > > > theory > > > > > of > > > > > what it does, but I've never had much luck controlling it. > > > > > > > > > > I need to read up and find how to tell apparmor to stop > > > > > controlling > > > > >
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
I can only remain on a version for so long. Eventually, updates have to happen, or an OS reinstalled, or something. I'd only be delaying the inevitable. Plus, I've updated a number of times now I've forgotten the last version that worked! Time to call it quits. Michelle. On Sun, 2023-09-24 at 21:27 +0200, Goetz T. Fischer wrote: > i didn't quite get the reason why you wanna keep using samba instead > of nfs but that aside, since this > is your home server, why not just keeping the indiana version that > worked? or at least the samba version > that worked? > security usually isn't that big of an issue for this purpose. > > On Sun, 24 Sep 2023 15:34:32 +0100, Michelle wrote: > > I would like to give my thanks to all who have run and maintained > > OpenIndiana for the many years it has been available. > > > > I have been using it since big red took the reins of Sun and > > changed > > the license of Open Solaris, as it gave me access to ZFS; a product > > that has made a big difference in my personal computing life. > > > > The community has been welcoming and patient with me despite my > > novice > > questions. I just wish I had the skills and time to be able to > > contribute to the project but my job, my caring duties and looking > > after a home have left me with little personal time... especially > > in > > this last few years. > > > > Unfortunately this latest issue with CIFS/SMB has been quite a > > disruption and with little way to make progress, I’ve determined > > that > > as the problems have persisted despite changing the client from > > Mint to > > Ubuntu Cinnamon, I now have to change server platform. I’ll be > > replacing my OI servers with Debian over the coming weeks. > > > > Thank you all for your patience and help over the years. > > > > Michelle. > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
I just updated all packages and tested again... Sep 24 20:31:37 mich-desktop kernel: [311047.112078] CIFS: VFS: \\192.168.0.2 RFC 1002 unknown response type 0xfe Sep 24 20:31:37 mich-desktop kernel: [311047.112114] CIFS: VFS: Push locks rc = -22 I'm at the end. Nowhere else to go. Michelle. On Sun, 2023-09-24 at 20:45 +0300, Toomas Soome via openindiana-discuss wrote: > Ou that is sad that hear. Btw, what is the oi build you are using on > that server? > > Sent from my iPhone > > > On 24. Sep 2023, at 17:34, Michelle wrote: > > > > I would like to give my thanks to all who have run and maintained > > OpenIndiana for the many years it has been available. > > > > I have been using it since big red took the reins of Sun and > > changed > > the license of Open Solaris, as it gave me access to ZFS; a product > > that has made a big difference in my personal computing life. > > > > The community has been welcoming and patient with me despite my > > novice > > questions. I just wish I had the skills and time to be able to > > contribute to the project but my job, my caring duties and looking > > after a home have left me with little personal time... especially > > in > > this last few years. > > > > Unfortunately this latest issue with CIFS/SMB has been quite a > > disruption and with little way to make progress, I’ve determined > > that > > as the problems have persisted despite changing the client from > > Mint to > > Ubuntu Cinnamon, I now have to change server platform. I’ll be > > replacing my OI servers with Debian over the coming weeks. > > > > Thank you all for your patience and help over the years. > > > > Michelle. > > > > > On Tue, 2023-09-19 at 20:54 +0100, Michelle wrote: > > > OK, I've now upgraded the client to Ubuntu Cinnamon 22.04 > > > > > > Accessing the text file on the OI share via BlueFish this time, > > > and > > > making changes and saving, I get (in syslog)... > > > > > > Sep 19 20:49:47 mich-desktop kernel: [13839.060593] CIFS: VFS: > > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > > Sep 19 20:49:47 mich-desktop kernel: [13839.060714] CIFS: VFS: > > > Push > > > locks rc = -22 > > > > > > Same behaviour as before when using xed on Mint. If I delete and > > > then > > > save, it works. And if I attempt to rename a file that's in use > > > (eg. > > > a > > > video) then I get resource unavailable. > > > > > > I have no realistic option now other than to conclude that it's > > > the > > > OpenIndiana side of things. > > > > > > I'd be grateful for guidance on whatever I need to file a bug > > > report > > > please. > > > > > > > > > > > > > On Mon, 2023-08-07 at 22:54 +0100, Michelle wrote: > > > > Apparmor is, indeed running, so I created a profile and put it > > > > to > > > > complain mode... noting that Libre Office is also in complain > > > > mode > > > > already. > > > > > > > > However, xed still fails with the same issue. It is, > > > > apparently, > > > > writing a temporary file and presumably failing... the same way > > > > that > > > > copying a file from the local machine to the share is > > > > failing... > > > > > > > > ... copying the file to a temporary .gooutputstream file but > > > > then > > > > somehow failing to delete the file that is about to be > > > > replaced. > > > > > > > > Michelle. > > > > > > > > On Mon, 2023-08-07 at 22:24 +0100, Michelle wrote: > > > > > I'm starting to believe that upgrading Linux has installed > > > > > apparmor > > > > > ... > > > > > and it's likely that apparmor is now getting in the way of > > > > > various > > > > > applications. > > > > > > > > > > I hate apparmor. I mean, I understand why it's there and the > > > > > theory > > > > > of > > > > > what it does, but I've never had much luck controlling it. > > > > > > > > > > I need to read up and find how to tell apparmor to stop > > > > > controlling > > > > > xed > > > > > and nemo, and see if that works. > > > > > > > &g
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
Hipster 2023.05 On Sun, 2023-09-24 at 20:45 +0300, Toomas Soome via openindiana-discuss wrote: > Ou that is sad that hear. Btw, what is the oi build you are using on > that server? > > Sent from my iPhone > > > On 24. Sep 2023, at 17:34, Michelle wrote: > > > > I would like to give my thanks to all who have run and maintained > > OpenIndiana for the many years it has been available. > > > > I have been using it since big red took the reins of Sun and > > changed > > the license of Open Solaris, as it gave me access to ZFS; a product > > that has made a big difference in my personal computing life. > > > > The community has been welcoming and patient with me despite my > > novice > > questions. I just wish I had the skills and time to be able to > > contribute to the project but my job, my caring duties and looking > > after a home have left me with little personal time... especially > > in > > this last few years. > > > > Unfortunately this latest issue with CIFS/SMB has been quite a > > disruption and with little way to make progress, I’ve determined > > that > > as the problems have persisted despite changing the client from > > Mint to > > Ubuntu Cinnamon, I now have to change server platform. I’ll be > > replacing my OI servers with Debian over the coming weeks. > > > > Thank you all for your patience and help over the years. > > > > Michelle. > > > > > On Tue, 2023-09-19 at 20:54 +0100, Michelle wrote: > > > OK, I've now upgraded the client to Ubuntu Cinnamon 22.04 > > > > > > Accessing the text file on the OI share via BlueFish this time, > > > and > > > making changes and saving, I get (in syslog)... > > > > > > Sep 19 20:49:47 mich-desktop kernel: [13839.060593] CIFS: VFS: > > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > > Sep 19 20:49:47 mich-desktop kernel: [13839.060714] CIFS: VFS: > > > Push > > > locks rc = -22 > > > > > > Same behaviour as before when using xed on Mint. If I delete and > > > then > > > save, it works. And if I attempt to rename a file that's in use > > > (eg. > > > a > > > video) then I get resource unavailable. > > > > > > I have no realistic option now other than to conclude that it's > > > the > > > OpenIndiana side of things. > > > > > > I'd be grateful for guidance on whatever I need to file a bug > > > report > > > please. > > > > > > > > > > > > > On Mon, 2023-08-07 at 22:54 +0100, Michelle wrote: > > > > Apparmor is, indeed running, so I created a profile and put it > > > > to > > > > complain mode... noting that Libre Office is also in complain > > > > mode > > > > already. > > > > > > > > However, xed still fails with the same issue. It is, > > > > apparently, > > > > writing a temporary file and presumably failing... the same way > > > > that > > > > copying a file from the local machine to the share is > > > > failing... > > > > > > > > ... copying the file to a temporary .gooutputstream file but > > > > then > > > > somehow failing to delete the file that is about to be > > > > replaced. > > > > > > > > Michelle. > > > > > > > > On Mon, 2023-08-07 at 22:24 +0100, Michelle wrote: > > > > > I'm starting to believe that upgrading Linux has installed > > > > > apparmor > > > > > ... > > > > > and it's likely that apparmor is now getting in the way of > > > > > various > > > > > applications. > > > > > > > > > > I hate apparmor. I mean, I understand why it's there and the > > > > > theory > > > > > of > > > > > what it does, but I've never had much luck controlling it. > > > > > > > > > > I need to read up and find how to tell apparmor to stop > > > > > controlling > > > > > xed > > > > > and nemo, and see if that works. > > > > > > > > > > Michelle. > > > > > > > > > > On Mon, 2023-08-07 at 22:20 +0100, Michelle wrote: > > > > > > I've translated up to the same cifs-utils version up to > > > > > > Mint > > > > &g
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
I would like to give my thanks to all who have run and maintained OpenIndiana for the many years it has been available. I have been using it since big red took the reins of Sun and changed the license of Open Solaris, as it gave me access to ZFS; a product that has made a big difference in my personal computing life. The community has been welcoming and patient with me despite my novice questions. I just wish I had the skills and time to be able to contribute to the project but my job, my caring duties and looking after a home have left me with little personal time... especially in this last few years. Unfortunately this latest issue with CIFS/SMB has been quite a disruption and with little way to make progress, I’ve determined that as the problems have persisted despite changing the client from Mint to Ubuntu Cinnamon, I now have to change server platform. I’ll be replacing my OI servers with Debian over the coming weeks. Thank you all for your patience and help over the years. Michelle. On Tue, 2023-09-19 at 20:54 +0100, Michelle wrote: > OK, I've now upgraded the client to Ubuntu Cinnamon 22.04 > > Accessing the text file on the OI share via BlueFish this time, and > making changes and saving, I get (in syslog)... > > Sep 19 20:49:47 mich-desktop kernel: [13839.060593] CIFS: VFS: > \\192.168.0.2 RFC 1002 unknown response type 0xfe > Sep 19 20:49:47 mich-desktop kernel: [13839.060714] CIFS: VFS: Push > locks rc = -22 > > Same behaviour as before when using xed on Mint. If I delete and then > save, it works. And if I attempt to rename a file that's in use (eg. > a > video) then I get resource unavailable. > > I have no realistic option now other than to conclude that it's the > OpenIndiana side of things. > > I'd be grateful for guidance on whatever I need to file a bug report > please. > > > > On Mon, 2023-08-07 at 22:54 +0100, Michelle wrote: > > Apparmor is, indeed running, so I created a profile and put it to > > complain mode... noting that Libre Office is also in complain mode > > already. > > > > However, xed still fails with the same issue. It is, apparently, > > writing a temporary file and presumably failing... the same way > > that > > copying a file from the local machine to the share is failing... > > > > ... copying the file to a temporary .gooutputstream file but then > > somehow failing to delete the file that is about to be replaced. > > > > Michelle. > > > > On Mon, 2023-08-07 at 22:24 +0100, Michelle wrote: > > > I'm starting to believe that upgrading Linux has installed > > > apparmor > > > ... > > > and it's likely that apparmor is now getting in the way of > > > various > > > applications. > > > > > > I hate apparmor. I mean, I understand why it's there and the > > > theory > > > of > > > what it does, but I've never had much luck controlling it. > > > > > > I need to read up and find how to tell apparmor to stop > > > controlling > > > xed > > > and nemo, and see if that works. > > > > > > Michelle. > > > > > > On Mon, 2023-08-07 at 22:20 +0100, Michelle wrote: > > > > I've translated up to the same cifs-utils version up to Mint > > > > 21.2 > > > > Victoria kernel 6.2.0-26 which is as far as I can go. I have > > > > also > > > > turned on nbmand on, on the servers. > > > > > > > > Oddly enough, Libre Office Writer can open and save the file on > > > > the > > > > share. "Text Editor" otherwise known as xed 3.4.3 ... can't. It > > > > reports > > > > the resource unavailable error. > > > > > > > > Copying a file from network share to local machine using Nemo > > > > 5.8.4 > > > > successfully copies the file, but the progress bar stops at > > > > 50%. > > > > I've > > > > still got the reports of the push locks and ... > > > > Aug 7 22:14:23 main-desktop kernel: [ 5181.959778] CIFS: VFS: > > > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > > > Aug 7 22:14:23 main-desktop kernel: [ 5181.959819] CIFS: VFS: > > > > Push > > > > locks rc = -22 > > > > ... on the client. > > > > > > > > Copying from the client to the server using the same version of > > > > Nemo, > > > > results in failure to overwrite the file with... > > > > Aug 7 22:17:09 main-desktop kernel: [ 5348.165575] CIFS: VFS: > > > >
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
OK, I've now upgraded the client to Ubuntu Cinnamon 22.04 Accessing the text file on the OI share via BlueFish this time, and making changes and saving, I get (in syslog)... Sep 19 20:49:47 mich-desktop kernel: [13839.060593] CIFS: VFS: \\192.168.0.2 RFC 1002 unknown response type 0xfe Sep 19 20:49:47 mich-desktop kernel: [13839.060714] CIFS: VFS: Push locks rc = -22 Same behaviour as before when using xed on Mint. If I delete and then save, it works. And if I attempt to rename a file that's in use (eg. a video) then I get resource unavailable. I have no realistic option now other than to conclude that it's the OpenIndiana side of things. I'd be grateful for guidance on whatever I need to file a bug report please. On Mon, 2023-08-07 at 22:54 +0100, Michelle wrote: > Apparmor is, indeed running, so I created a profile and put it to > complain mode... noting that Libre Office is also in complain mode > already. > > However, xed still fails with the same issue. It is, apparently, > writing a temporary file and presumably failing... the same way that > copying a file from the local machine to the share is failing... > > ... copying the file to a temporary .gooutputstream file but then > somehow failing to delete the file that is about to be replaced. > > Michelle. > > On Mon, 2023-08-07 at 22:24 +0100, Michelle wrote: > > I'm starting to believe that upgrading Linux has installed apparmor > > ... > > and it's likely that apparmor is now getting in the way of various > > applications. > > > > I hate apparmor. I mean, I understand why it's there and the theory > > of > > what it does, but I've never had much luck controlling it. > > > > I need to read up and find how to tell apparmor to stop controlling > > xed > > and nemo, and see if that works. > > > > Michelle. > > > > On Mon, 2023-08-07 at 22:20 +0100, Michelle wrote: > > > I've translated up to the same cifs-utils version up to Mint 21.2 > > > Victoria kernel 6.2.0-26 which is as far as I can go. I have also > > > turned on nbmand on, on the servers. > > > > > > Oddly enough, Libre Office Writer can open and save the file on > > > the > > > share. "Text Editor" otherwise known as xed 3.4.3 ... can't. It > > > reports > > > the resource unavailable error. > > > > > > Copying a file from network share to local machine using Nemo > > > 5.8.4 > > > successfully copies the file, but the progress bar stops at 50%. > > > I've > > > still got the reports of the push locks and ... > > > Aug 7 22:14:23 main-desktop kernel: [ 5181.959778] CIFS: VFS: > > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > > Aug 7 22:14:23 main-desktop kernel: [ 5181.959819] CIFS: VFS: > > > Push > > > locks rc = -22 > > > ... on the client. > > > > > > Copying from the client to the server using the same version of > > > Nemo, > > > results in failure to overwrite the file with... > > > Aug 7 22:17:09 main-desktop kernel: [ 5348.165575] CIFS: VFS: > > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > > Aug 7 22:17:09 main-desktop kernel: [ 5348.165613] CIFS: VFS: > > > Push > > > locks rc = -22 > > > ... although deleting the file and then copying it up to the > > > share, > > > works. Again, it looks like there's a problem deleting the > > > original > > > file in order to rename the temporary file to the target file > > > name. > > > > > > Michelle. > > > > > > On Mon, 2023-08-07 at 19:35 +0300, Toomas Soome via openindiana- > > > discuss > > > wrote: > > > > hi! > > > > > > > > I just did test with ubuntu 22.04.01, cifs-utils: > > > > 2:6.14-1ubuntu0.1 > > > > > > > > And seems to behave as expected. My share has nbmand=on, > > > > however. > > > > > > > > rgds, > > > > toomas > > > > > > > > > On 7. Aug 2023, at 19:07, Michelle > > > > > wrote: > > > > > > > > > > After that updating, I now have a problem copying files to > > > > > and > > > > > from > > > > > the > > > > > share and the client. > > > > > > > > > > Firstly... using command line to go onto the share and use > > > > > Nano > > > > > to > > > > > edit > > > > >
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
Apparmor is, indeed running, so I created a profile and put it to complain mode... noting that Libre Office is also in complain mode already. However, xed still fails with the same issue. It is, apparently, writing a temporary file and presumably failing... the same way that copying a file from the local machine to the share is failing... ... copying the file to a temporary .gooutputstream file but then somehow failing to delete the file that is about to be replaced. Michelle. On Mon, 2023-08-07 at 22:24 +0100, Michelle wrote: > I'm starting to believe that upgrading Linux has installed apparmor > ... > and it's likely that apparmor is now getting in the way of various > applications. > > I hate apparmor. I mean, I understand why it's there and the theory > of > what it does, but I've never had much luck controlling it. > > I need to read up and find how to tell apparmor to stop controlling > xed > and nemo, and see if that works. > > Michelle. > > On Mon, 2023-08-07 at 22:20 +0100, Michelle wrote: > > I've translated up to the same cifs-utils version up to Mint 21.2 > > Victoria kernel 6.2.0-26 which is as far as I can go. I have also > > turned on nbmand on, on the servers. > > > > Oddly enough, Libre Office Writer can open and save the file on the > > share. "Text Editor" otherwise known as xed 3.4.3 ... can't. It > > reports > > the resource unavailable error. > > > > Copying a file from network share to local machine using Nemo 5.8.4 > > successfully copies the file, but the progress bar stops at 50%. > > I've > > still got the reports of the push locks and ... > > Aug 7 22:14:23 main-desktop kernel: [ 5181.959778] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 22:14:23 main-desktop kernel: [ 5181.959819] CIFS: VFS: Push > > locks rc = -22 > > ... on the client. > > > > Copying from the client to the server using the same version of > > Nemo, > > results in failure to overwrite the file with... > > Aug 7 22:17:09 main-desktop kernel: [ 5348.165575] CIFS: VFS: > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > Aug 7 22:17:09 main-desktop kernel: [ 5348.165613] CIFS: VFS: Push > > locks rc = -22 > > ... although deleting the file and then copying it up to the share, > > works. Again, it looks like there's a problem deleting the original > > file in order to rename the temporary file to the target file name. > > > > Michelle. > > > > On Mon, 2023-08-07 at 19:35 +0300, Toomas Soome via openindiana- > > discuss > > wrote: > > > hi! > > > > > > I just did test with ubuntu 22.04.01, cifs-utils: > > > 2:6.14-1ubuntu0.1 > > > > > > And seems to behave as expected. My share has nbmand=on, however. > > > > > > rgds, > > > toomas > > > > > > > On 7. Aug 2023, at 19:07, Michelle > > > > wrote: > > > > > > > > After that updating, I now have a problem copying files to and > > > > from > > > > the > > > > share and the client. > > > > > > > > Firstly... using command line to go onto the share and use Nano > > > > to > > > > edit > > > > a text file... no problem. > > > > > > > > Using Nano to browse to the mounted share, open in Text Editor, > > > > edit > > > > and then save... it gives "resource unavailable." > > > > > > > > When using Nemo to copy from the share to the local Linux > > > > machine, > > > > it > > > > copies the small test file successfully, but hangs on the > > > > "progress > > > > bar" which never closes. I have to kill Nemo. > > > > > > > > When using Nemo to copy from the local Linux machine to the > > > > server > > > > share, it copies up the file using the temp name, but can't > > > > actually > > > > replace the file. > > > > > > > > If I manually delete the file from the server, and then copy > > > > the > > > > file > > > > up to the share... it works. > > > > > > > > All the time, this... > > > > Aug 7 16:54:16 main-desktop kernel: [ 772.668870] CIFS: VFS: > > > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > > > Aug 7 16:54:16 main-desktop kernel: [ 772.668915] CIFS: VFS: > > > > Push > > > > locks rc = -22 >
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
I'm starting to believe that upgrading Linux has installed apparmor ... and it's likely that apparmor is now getting in the way of various applications. I hate apparmor. I mean, I understand why it's there and the theory of what it does, but I've never had much luck controlling it. I need to read up and find how to tell apparmor to stop controlling xed and nemo, and see if that works. Michelle. On Mon, 2023-08-07 at 22:20 +0100, Michelle wrote: > I've translated up to the same cifs-utils version up to Mint 21.2 > Victoria kernel 6.2.0-26 which is as far as I can go. I have also > turned on nbmand on, on the servers. > > Oddly enough, Libre Office Writer can open and save the file on the > share. "Text Editor" otherwise known as xed 3.4.3 ... can't. It > reports > the resource unavailable error. > > Copying a file from network share to local machine using Nemo 5.8.4 > successfully copies the file, but the progress bar stops at 50%. I've > still got the reports of the push locks and ... > Aug 7 22:14:23 main-desktop kernel: [ 5181.959778] CIFS: VFS: > \\192.168.0.4 RFC 1002 unknown response type 0xfe > Aug 7 22:14:23 main-desktop kernel: [ 5181.959819] CIFS: VFS: Push > locks rc = -22 > ... on the client. > > Copying from the client to the server using the same version of Nemo, > results in failure to overwrite the file with... > Aug 7 22:17:09 main-desktop kernel: [ 5348.165575] CIFS: VFS: > \\192.168.0.2 RFC 1002 unknown response type 0xfe > Aug 7 22:17:09 main-desktop kernel: [ 5348.165613] CIFS: VFS: Push > locks rc = -22 > ... although deleting the file and then copying it up to the share, > works. Again, it looks like there's a problem deleting the original > file in order to rename the temporary file to the target file name. > > Michelle. > > On Mon, 2023-08-07 at 19:35 +0300, Toomas Soome via openindiana- > discuss > wrote: > > hi! > > > > I just did test with ubuntu 22.04.01, cifs-utils: > > 2:6.14-1ubuntu0.1 > > > > And seems to behave as expected. My share has nbmand=on, however. > > > > rgds, > > toomas > > > > > On 7. Aug 2023, at 19:07, Michelle wrote: > > > > > > After that updating, I now have a problem copying files to and > > > from > > > the > > > share and the client. > > > > > > Firstly... using command line to go onto the share and use Nano > > > to > > > edit > > > a text file... no problem. > > > > > > Using Nano to browse to the mounted share, open in Text Editor, > > > edit > > > and then save... it gives "resource unavailable." > > > > > > When using Nemo to copy from the share to the local Linux > > > machine, > > > it > > > copies the small test file successfully, but hangs on the > > > "progress > > > bar" which never closes. I have to kill Nemo. > > > > > > When using Nemo to copy from the local Linux machine to the > > > server > > > share, it copies up the file using the temp name, but can't > > > actually > > > replace the file. > > > > > > If I manually delete the file from the server, and then copy the > > > file > > > up to the share... it works. > > > > > > All the time, this... > > > Aug 7 16:54:16 main-desktop kernel: [ 772.668870] CIFS: VFS: > > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > > Aug 7 16:54:16 main-desktop kernel: [ 772.668915] CIFS: VFS: > > > Push > > > locks rc = -22 > > > ... is prevalent in the client side syslog when this is > > > happening. > > > > > > There was a previous post in mid 2022 about a bug on Linux and > > > that > > > the > > > workaround was to force vers=2.0 but I've tried that and there's > > > no > > > change in the behaviour. > > > > > > From what I've read, my nbmand property is best set to off, which > > > is > > > what it is. Or do I have that wrong? > > > > > > I'm now at a loss as to where to continue looking. > > > > > > OpenIndiana Hipster 2023.05 (powered by illumos) > > > > > > Distributor ID: Linuxmint > > > Description:Linux Mint 21 > > > Release:21 > > > Codename: vanessa > > > > > > mount from util-linux 2.37.2 (libmount 2.37.2: selinux, smack, > > > btrfs, > > > verity, namespaces, assert, debug) > > > > > &
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
I've translated up to the same cifs-utils version up to Mint 21.2 Victoria kernel 6.2.0-26 which is as far as I can go. I have also turned on nbmand on, on the servers. Oddly enough, Libre Office Writer can open and save the file on the share. "Text Editor" otherwise known as xed 3.4.3 ... can't. It reports the resource unavailable error. Copying a file from network share to local machine using Nemo 5.8.4 successfully copies the file, but the progress bar stops at 50%. I've still got the reports of the push locks and ... Aug 7 22:14:23 main-desktop kernel: [ 5181.959778] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 22:14:23 main-desktop kernel: [ 5181.959819] CIFS: VFS: Push locks rc = -22 ... on the client. Copying from the client to the server using the same version of Nemo, results in failure to overwrite the file with... Aug 7 22:17:09 main-desktop kernel: [ 5348.165575] CIFS: VFS: \\192.168.0.2 RFC 1002 unknown response type 0xfe Aug 7 22:17:09 main-desktop kernel: [ 5348.165613] CIFS: VFS: Push locks rc = -22 ... although deleting the file and then copying it up to the share, works. Again, it looks like there's a problem deleting the original file in order to rename the temporary file to the target file name. Michelle. On Mon, 2023-08-07 at 19:35 +0300, Toomas Soome via openindiana-discuss wrote: > hi! > > I just did test with ubuntu 22.04.01, cifs-utils: > 2:6.14-1ubuntu0.1 > > And seems to behave as expected. My share has nbmand=on, however. > > rgds, > toomas > > > On 7. Aug 2023, at 19:07, Michelle wrote: > > > > After that updating, I now have a problem copying files to and from > > the > > share and the client. > > > > Firstly... using command line to go onto the share and use Nano to > > edit > > a text file... no problem. > > > > Using Nano to browse to the mounted share, open in Text Editor, > > edit > > and then save... it gives "resource unavailable." > > > > When using Nemo to copy from the share to the local Linux machine, > > it > > copies the small test file successfully, but hangs on the "progress > > bar" which never closes. I have to kill Nemo. > > > > When using Nemo to copy from the local Linux machine to the server > > share, it copies up the file using the temp name, but can't > > actually > > replace the file. > > > > If I manually delete the file from the server, and then copy the > > file > > up to the share... it works. > > > > All the time, this... > > Aug 7 16:54:16 main-desktop kernel: [ 772.668870] CIFS: VFS: > > \\192.168.0.2 RFC 1002 unknown response type 0xfe > > Aug 7 16:54:16 main-desktop kernel: [ 772.668915] CIFS: VFS: Push > > locks rc = -22 > > ... is prevalent in the client side syslog when this is happening. > > > > There was a previous post in mid 2022 about a bug on Linux and that > > the > > workaround was to force vers=2.0 but I've tried that and there's no > > change in the behaviour. > > > > From what I've read, my nbmand property is best set to off, which > > is > > what it is. Or do I have that wrong? > > > > I'm now at a loss as to where to continue looking. > > > > OpenIndiana Hipster 2023.05 (powered by illumos) > > > > Distributor ID: Linuxmint > > Description:Linux Mint 21 > > Release: 21 > > Codename: vanessa > > > > mount from util-linux 2.37.2 (libmount 2.37.2: selinux, smack, > > btrfs, > > verity, namespaces, assert, debug) > > > > > > > > On Mon, 2023-08-07 at 10:08 +0100, Michelle wrote: > > > nbmand is off on both servers. > > > > > > Michelle. > > > > > > On Mon, 2023-08-07 at 11:54 +0300, Toomas Soome wrote: > > > > > > > > > > > > > On 7. Aug 2023, at 11:38, Michelle > > > > > wrote: > > > > > > > > > > I did a quick test after updating and rebooting both servers. > > > > > > > > > > Copying files from one of the servers (panther - 0.4) to > > > > > another > > > > > server > > > > > (jaguar - 0.2) via the Linux client ... resulted in the > > > > > following > > > > > in > > > > > the log on the client, however, there was nothing in the > > > > > syslog > > > > > on > > > > > either of the servers. > > > > > > > > > > There does seem to be a mention of a patch in Linux SMB as of
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
After that updating, I now have a problem copying files to and from the share and the client. Firstly... using command line to go onto the share and use Nano to edit a text file... no problem. Using Nano to browse to the mounted share, open in Text Editor, edit and then save... it gives "resource unavailable." When using Nemo to copy from the share to the local Linux machine, it copies the small test file successfully, but hangs on the "progress bar" which never closes. I have to kill Nemo. When using Nemo to copy from the local Linux machine to the server share, it copies up the file using the temp name, but can't actually replace the file. If I manually delete the file from the server, and then copy the file up to the share... it works. All the time, this... Aug 7 16:54:16 main-desktop kernel: [ 772.668870] CIFS: VFS: \\192.168.0.2 RFC 1002 unknown response type 0xfe Aug 7 16:54:16 main-desktop kernel: [ 772.668915] CIFS: VFS: Push locks rc = -22 ... is prevalent in the client side syslog when this is happening. There was a previous post in mid 2022 about a bug on Linux and that the workaround was to force vers=2.0 but I've tried that and there's no change in the behaviour. From what I've read, my nbmand property is best set to off, which is what it is. Or do I have that wrong? I'm now at a loss as to where to continue looking. OpenIndiana Hipster 2023.05 (powered by illumos) Distributor ID: Linuxmint Description:Linux Mint 21 Release:21 Codename: vanessa mount from util-linux 2.37.2 (libmount 2.37.2: selinux, smack, btrfs, verity, namespaces, assert, debug) On Mon, 2023-08-07 at 10:08 +0100, Michelle wrote: > nbmand is off on both servers. > > Michelle. > > On Mon, 2023-08-07 at 11:54 +0300, Toomas Soome wrote: > > > > > > > On 7. Aug 2023, at 11:38, Michelle wrote: > > > > > > I did a quick test after updating and rebooting both servers. > > > > > > Copying files from one of the servers (panther - 0.4) to another > > > server > > > (jaguar - 0.2) via the Linux client ... resulted in the following > > > in > > > the log on the client, however, there was nothing in the syslog > > > on > > > either of the servers. > > > > > > There does seem to be a mention of a patch in Linux SMB as of > > > late > > > June > > > this year, but having said that, there seem to be a number of > > > patches > > > throughout the years regarding this, but I'm unable to get a grip > > > with > > > it. > > > > > > Aug 7 09:28:35 main-desktop systemd[1109]: Starting GNOME > > > Terminal > > > Server... > > > Aug 7 09:28:35 main-desktop dbus-daemon[1136]: [session uid=1101 > > > pid=1136] Successfully activated service 'org.gnome.Terminal' > > > Aug 7 09:28:35 main-desktop systemd[1109]: Started GNOME > > > Terminal > > > Server. > > > Aug 7 09:28:35 main-desktop systemd[1109]: Started VTE child > > > process > > > 1644922 launched by gnome-terminal-server process 1644897. > > > Aug 7 09:30:01 main-desktop CRON[1645001]: (root) CMD ([ -x > > > /etc/init.d/anacron ] && if [ ! -d /run/systemd/system ]; then > > > /usr/sbin/invoke-rc.d anacron start >/dev/null; fi) > > > Aug 7 09:30:10 main-desktop kernel: [2945189.248744] CIFS: VFS: > > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > > > I’m not really sure where is this coming from, probably want to > > check > > netowk packet with tshark/wireshark. > > > > > Aug 7 09:30:10 main-desktop kernel: [2945189.248788] CIFS: VFS: > > > Push > > > locks rc = -22 > > > > Do you have zfs nbmand property set on? > > > > rgds, > > toomas > > > > > Aug 7 09:30:20 main-desktop kernel: [2945198.711222] CIFS: VFS: > > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > > Aug 7 09:30:25 main-desktop kernel: [2945204.406488] CIFS: VFS: > > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > > Aug 7 09:30:28 main-desktop kernel: [2945206.802407] CIFS: VFS: > > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > > Aug 7 09:30:30 main-desktop kernel: [2945208.649958] CIFS: VFS: > > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > > Aug 7 09:30:32 main-desktop kernel: [2945210.907929] CIFS: VFS: > > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > > Aug 7 09:30:33 main-desktop kernel: [2945212.056828] CIFS: VFS: > > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > > Au
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
nbmand is off on both servers. Michelle. On Mon, 2023-08-07 at 11:54 +0300, Toomas Soome wrote: > > > > On 7. Aug 2023, at 11:38, Michelle wrote: > > > > I did a quick test after updating and rebooting both servers. > > > > Copying files from one of the servers (panther - 0.4) to another > > server > > (jaguar - 0.2) via the Linux client ... resulted in the following > > in > > the log on the client, however, there was nothing in the syslog on > > either of the servers. > > > > There does seem to be a mention of a patch in Linux SMB as of late > > June > > this year, but having said that, there seem to be a number of > > patches > > throughout the years regarding this, but I'm unable to get a grip > > with > > it. > > > > Aug 7 09:28:35 main-desktop systemd[1109]: Starting GNOME Terminal > > Server... > > Aug 7 09:28:35 main-desktop dbus-daemon[1136]: [session uid=1101 > > pid=1136] Successfully activated service 'org.gnome.Terminal' > > Aug 7 09:28:35 main-desktop systemd[1109]: Started GNOME Terminal > > Server. > > Aug 7 09:28:35 main-desktop systemd[1109]: Started VTE child > > process > > 1644922 launched by gnome-terminal-server process 1644897. > > Aug 7 09:30:01 main-desktop CRON[1645001]: (root) CMD ([ -x > > /etc/init.d/anacron ] && if [ ! -d /run/systemd/system ]; then > > /usr/sbin/invoke-rc.d anacron start >/dev/null; fi) > > Aug 7 09:30:10 main-desktop kernel: [2945189.248744] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > I’m not really sure where is this coming from, probably want to check > netowk packet with tshark/wireshark. > > > Aug 7 09:30:10 main-desktop kernel: [2945189.248788] CIFS: VFS: > > Push > > locks rc = -22 > > Do you have zfs nbmand property set on? > > rgds, > toomas > > > Aug 7 09:30:20 main-desktop kernel: [2945198.711222] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:25 main-desktop kernel: [2945204.406488] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:28 main-desktop kernel: [2945206.802407] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:30 main-desktop kernel: [2945208.649958] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:32 main-desktop kernel: [2945210.907929] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:33 main-desktop kernel: [2945212.056828] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:36 main-desktop kernel: [2945214.981583] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:37 main-desktop kernel: [2945216.163547] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:37 main-desktop systemd[1109]: Started VTE child > > process > > 1645538 launched by gnome-terminal-server process 1644897. > > Aug 7 09:30:39 main-desktop kernel: [2945218.019954] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:41 main-desktop kernel: [2945219.860681] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:41 main-desktop kernel: [2945220.282600] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:41 main-desktop kernel: [2945220.282649] CIFS: VFS: > > Push > > locks rc = -22 > > Aug 7 09:30:44 main-desktop kernel: [2945222.937137] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:44 main-desktop kernel: [2945222.937168] CIFS: VFS: > > Push > > locks rc = -22 > > Aug 7 09:30:48 main-desktop kernel: [2945227.269625] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:54 main-desktop kernel: [2945233.354499] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:30:57 main-desktop kernel: [2945235.676076] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:31:02 main-desktop kernel: [2945241.140574] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:31:06 main-desktop kernel: [2945244.760306] CIFS: VFS: > > \\192.168.0.4 RFC 1002 unknown response type 0xfe > > Aug 7 09:31:06 main-desktop systemd[1]: Started Run anacron jobs. > > Aug 7 09:31:06 main-desktop anacron[1645804]: Anacron 2.3 started > > on > > 2023-08-07 > > Aug 7 09:31:06 main-desktop anacron[1645804]: N
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
I did a quick test after updating and rebooting both servers. Copying files from one of the servers (panther - 0.4) to another server (jaguar - 0.2) via the Linux client ... resulted in the following in the log on the client, however, there was nothing in the syslog on either of the servers. There does seem to be a mention of a patch in Linux SMB as of late June this year, but having said that, there seem to be a number of patches throughout the years regarding this, but I'm unable to get a grip with it. Aug 7 09:28:35 main-desktop systemd[1109]: Starting GNOME Terminal Server... Aug 7 09:28:35 main-desktop dbus-daemon[1136]: [session uid=1101 pid=1136] Successfully activated service 'org.gnome.Terminal' Aug 7 09:28:35 main-desktop systemd[1109]: Started GNOME Terminal Server. Aug 7 09:28:35 main-desktop systemd[1109]: Started VTE child process 1644922 launched by gnome-terminal-server process 1644897. Aug 7 09:30:01 main-desktop CRON[1645001]: (root) CMD ([ -x /etc/init.d/anacron ] && if [ ! -d /run/systemd/system ]; then /usr/sbin/invoke-rc.d anacron start >/dev/null; fi) Aug 7 09:30:10 main-desktop kernel: [2945189.248744] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:10 main-desktop kernel: [2945189.248788] CIFS: VFS: Push locks rc = -22 Aug 7 09:30:20 main-desktop kernel: [2945198.711222] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:25 main-desktop kernel: [2945204.406488] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:28 main-desktop kernel: [2945206.802407] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:30 main-desktop kernel: [2945208.649958] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:32 main-desktop kernel: [2945210.907929] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:33 main-desktop kernel: [2945212.056828] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:36 main-desktop kernel: [2945214.981583] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:37 main-desktop kernel: [2945216.163547] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:37 main-desktop systemd[1109]: Started VTE child process 1645538 launched by gnome-terminal-server process 1644897. Aug 7 09:30:39 main-desktop kernel: [2945218.019954] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:41 main-desktop kernel: [2945219.860681] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:41 main-desktop kernel: [2945220.282600] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:41 main-desktop kernel: [2945220.282649] CIFS: VFS: Push locks rc = -22 Aug 7 09:30:44 main-desktop kernel: [2945222.937137] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:44 main-desktop kernel: [2945222.937168] CIFS: VFS: Push locks rc = -22 Aug 7 09:30:48 main-desktop kernel: [2945227.269625] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:54 main-desktop kernel: [2945233.354499] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:30:57 main-desktop kernel: [2945235.676076] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:31:02 main-desktop kernel: [2945241.140574] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:31:06 main-desktop kernel: [2945244.760306] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:31:06 main-desktop systemd[1]: Started Run anacron jobs. Aug 7 09:31:06 main-desktop anacron[1645804]: Anacron 2.3 started on 2023-08-07 Aug 7 09:31:06 main-desktop anacron[1645804]: Normal exit (0 jobs run) Aug 7 09:31:06 main-desktop systemd[1]: anacron.service: Deactivated successfully. Aug 7 09:31:08 main-desktop kernel: [2945247.431492] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:31:14 main-desktop kernel: [2945253.154480] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe Aug 7 09:31:14 main-desktop kernel: [2945253.255401] CIFS: VFS: \\192.168.0.4 RFC 1002 unknown response type 0xfe e On Mon, 2023-08-07 at 11:16 +0300, Toomas Soome wrote: > > > > On 7. Aug 2023, at 11:12, Michelle wrote: > > > > Hi Toomas, > > > > Thanks for the feedback. > > > > I'll update both servers and run more tests. > > > > Michelle. > > For SMB, uid/gid sync is not important, you will access the share > with authenticated user (or guest). > > With NFS auth sys mechanism, the usernames and uid/gid need to match, > and your nfsv4_domain should match with client system setting (or > your access will be mapped to user ’nobody’). > > rgds, > toomas > > > > > On Mon, 2023-08-07 at 10:56 +0300, Toomas Soome via openi
Re: [OpenIndiana-discuss] SMB NFS not sure where to go
Hi Toomas, Thanks for the feedback. I'll update both servers and run more tests. Michelle. On Mon, 2023-08-07 at 10:56 +0300, Toomas Soome via openindiana-discuss wrote: > > > > On 7. Aug 2023, at 09:21, Michelle wrote: > > > > OI server is at... > > OpenIndiana Hipster 2022.10 (powered by illumos) > > > Please run ‘pkg update’. We have had many updates since 2022. > > > > > Client is Linux Mint... > > Distributor ID: Linuxmint > > Description:Linux Mint 21 > > Release:21 > > Codename: vanessa > > > > I'm continuing to have problems with SMB shares. Even mounting with > > no > > cache... > > sudo mount //192.168.0.2/jaguar /mnt/jaguar -o > > username=michelle,password=password,file_mode=0777,dir_mode=0777,ca > > che= > > none > > ... > > //192.168.0.2/jaguar on /mnt/jaguar type cifs > > (rw,relatime,vers=3.0,cache=none,username=michelle,uid=0,noforceuid > > ,gid > > =0,noforcegid,addr=192.168.0.2,file_mode=0777,dir_mode=0777,soft,no > > unix > > ,mapposix,rsize=65536,wsize=65536,bsize=1048576,echo_interval=60,ac > > time > > o=1) > > > > (for info, the UID of michelle is the same on client and server) > > ...I'm getting, "resource unavailable," on various read- > > write/rename > > operations including deleting files. Say, sixty files, three will > > fail > > to delete giving "resource unavailable" and then I have to delete > > them > > again, and they'll go. > > > > When the Linux client hits resource unavaialble, there's nothing in > > the > > Linux messages and nothing in the OI /var/adm/messages either, that > > gives a hint of what's going on. > > > > Having hit a brick wall with CIFS, I'm trying to configure an NFS > > share > > against user rather than machine IP, but I'm getting nowhere. > > > > zfs set sharenfs='rw=.:michelle' panther > > access list used with rw does not allow to set per user access, see > share_nfs(8). It should interpret michelle as hostname, but I’m not > sure why it will complain about invalid option below. > > > cannot set property for 'panther': 'sharenfs' cannot be set to > > invalid > > options > > > > I'm tying myself in knots because it seems that there's "legacy" > > NFS, > > ZFS NFS with different options/restrictions and I'm not sure where > > to > > go. If I read correctly, only ZFS NFS is installed by "default" on > > a > > new OI installation, but I'm not about to start going installing > > stuff > > until I have a better idea what I'm doing. > > > > One reason I set some shares against user permissions is to protect > > against ransomware. I unmount and re-mount rw when I want to write > > to > > them, hence pure hostip is a problem for me. > > > > I don't know where I'm looking for either option. > > > > Hence the decision to look to NFS (which I've never really had to > > handle before) and I'm hitting trouble there. > > > > Grateful for thoughts and advice. > > > > Michelle. > > > > zfs sharenfs is just one possible way to implement sharing, it is > more useful when you are switching pools between different machines - > so you would get sharing options with pool and do not have to > transfer /etc/dfs/dfstab between the systems. > > For debugging, I suggest to test with command ’share’ first - make > sure your svc:/network/nfs/server:default is online, then use share > command to test the options for specific share. Once happy, you can > use either /etc/dfs/dfstab or zfs sharenfs (if you still do get the > error, please report the bug in https://www.illumos.org/issues/ > > And please do keep the system up to date, so we know what to expect - > we only do rolling updates and we are not patching old versions. > > rgds, > toomas > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] SMB NFS not sure where to go
OI server is at... OpenIndiana Hipster 2022.10 (powered by illumos) Client is Linux Mint... Distributor ID: Linuxmint Description:Linux Mint 21 Release:21 Codename: vanessa I'm continuing to have problems with SMB shares. Even mounting with no cache... sudo mount //192.168.0.2/jaguar /mnt/jaguar -o username=michelle,password=password,file_mode=0777,dir_mode=0777,cache= none ... //192.168.0.2/jaguar on /mnt/jaguar type cifs (rw,relatime,vers=3.0,cache=none,username=michelle,uid=0,noforceuid,gid =0,noforcegid,addr=192.168.0.2,file_mode=0777,dir_mode=0777,soft,nounix ,mapposix,rsize=65536,wsize=65536,bsize=1048576,echo_interval=60,actime o=1) (for info, the UID of michelle is the same on client and server) ...I'm getting, "resource unavailable," on various read-write/rename operations including deleting files. Say, sixty files, three will fail to delete giving "resource unavailable" and then I have to delete them again, and they'll go. When the Linux client hits resource unavaialble, there's nothing in the Linux messages and nothing in the OI /var/adm/messages either, that gives a hint of what's going on. Having hit a brick wall with CIFS, I'm trying to configure an NFS share against user rather than machine IP, but I'm getting nowhere. zfs set sharenfs='rw=.:michelle' panther cannot set property for 'panther': 'sharenfs' cannot be set to invalid options I'm tying myself in knots because it seems that there's "legacy" NFS, ZFS NFS with different options/restrictions and I'm not sure where to go. If I read correctly, only ZFS NFS is installed by "default" on a new OI installation, but I'm not about to start going installing stuff until I have a better idea what I'm doing. One reason I set some shares against user permissions is to protect against ransomware. I unmount and re-mount rw when I want to write to them, hence pure hostip is a problem for me. I don't know where I'm looking for either option. Hence the decision to look to NFS (which I've never really had to handle before) and I'm hitting trouble there. Grateful for thoughts and advice. Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Deleting empty directories
Thanks Tim. That's very useful learning info. I didn't know about the --prune- empty-dirs option. I guess I'm guilty of cribbing what I need from the internet and never learning things properly. Oh for the time to do all the things I would like to! Many thanks, Michelle. On Wed, 2023-08-02 at 16:32 -0500, Tim Mooney via openindiana-discuss wrote: > In regard to: [OpenIndiana-discuss] Deleting empty directories, > Michelle...: > > > Basically, I extract all the flac files out of my main music > > collection, to a separate location, (using rsync) and then I have > > to > > delete the empty directories that rsync has created. > > You could use '--prune-empty-dirs' with rsync. > > > I managed the first part, no problem. However, it's the deleting of > > empty directories that is causing me trouble. > > > > The Linux command is... > > find /mnt/jaguar/users/michelle/FlacCopy -type d -empty -delete > > ...however the OI version of find doesn't support -empty > > No, but it supports '-depth', which means you can > > find /mnt/jaguar/users/michelle/FlacCopy -depth -type d \ > -exec rmdir {} \; 2>/dev/null > > That will fail *a lot*, but the error messages are thrown away and > the > empty directories will be deleted. > > or if you have GNU fileutils installed (so 'grmdir' exists) > > find /mnt/jaguar/users/michelle/FlacCopy -depth -type d \ > -exec grmdir --ignore-fail-on-non-empty {} \; > > -depth is important for either of these to work. > > The best part of both of those is that you can insert 'echo' between > '-exec' and the start of the command, and it will show you what it's > going to do without doing it. If you like what you see, just remove > the > 'echo'. > > > I've tried to install gfind, but that doesn't appear to be > > available. > > Someone else already answered with the correct pkg FMRI. > > Tim ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Deleting empty directories
Many thanks Volker. That'll do the trick. Michelle. On Wed, 2023-08-02 at 23:19 +0200, Volker A. Brandt wrote: > Michelle writes: > [...] > > I've tried to install gfind, but that doesn't appear to be > > available. > > gfind is in the package "file/gnu-findutils". > > > Best regards -- Volker ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Deleting empty directories
At the moment I'm doing something on an OI server which is executed by a Linux client. However, due to the "Resource temporarily unavailable (11)" through the linux mounting (even with the cache off, etc.) I'm having to try and replicate this on the OI server itself. Basically, I extract all the flac files out of my main music collection, to a separate location, (using rsync) and then I have to delete the empty directories that rsync has created. I managed the first part, no problem. However, it's the deleting of empty directories that is causing me trouble. The Linux command is... find /mnt/jaguar/users/michelle/FlacCopy -type d -empty -delete ...however the OI version of find doesn't support -empty I've tried to install gfind, but that doesn't appear to be available. I've tried to follow various scripts... (the below is just to print empty directories at the moment, for testing) find /mnt/jaguar/users/michelle/FlacCopy -type d | xargs -i ls -ed {} | awk '{if ($5=="2"){ print $0 }}' ...but I fall foul that some of the names in the directories contain ' (eg. Sinead O'Connor) which then causes the command to fail with, "xargs: Missing quote" Any clues would be gratefully received. Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Readahead Batch error
It's a CIFS mount. Hard doesn't take, it always mounts as soft. Michelle. On Fri, 2023-07-21 at 21:37 +0200, Udo Grabowski (IMK) wrote: > On 21/07/2023 20:55, Michelle wrote: > > In case it helps, this is what's reported by mount... > > > > //192.168.0.2/jaguar on /mnt/jaguar type cifs > > (rw,relatime,vers=3.0,cache=none,username=michelle,.. > ..,soft,... > There's your problem: NEVER EVER mount anything > SOFT, but always HARD ! SOFT means YOU have to deal with > every error the channel delivers ! > > This option should gnererally be hidden from users in > the documentation, it should never be used normally. > > > > Michelle. > > > > On Fri, 2023-07-21 at 19:37 +0200, Till Wegmüller wrote: > > > Hey Michelle > > > > > > The Internet tells me this might be related to linuxes client > > > caching. > > > Try cache=none if you are having locking problems. And experiment > > > a > > > bit > > > with the cache variable. > > > > > > -Till > > > On 21.07.23 13:41, Michelle wrote: > > > > Can't work out why I'm getting this, which is causing problems > > > > accessing/renaming/moving files, more specifically if they're > > > > in > > > > use by > > > > another process, which used to be no problem under vers=1.0 > > > > > > > > ...but obviously I'm not going to shift back to vers=1.0. I did > > > > try > > > > connecting at 2.0 and also leaving the vers off the command > > > > (detailed > > > > later) but no change in behaviour. > > > > > > > > Looking the error up on the internet has not given anything > > > > that I > > > > can > > > > believe is relevant to this particular combination of client > > > > and > > > > server, etc. > > > > > > > > OI server is at... > > > > OpenIndiana Hipster 2022.10 (powered by illumos) > > > > OpenIndiana Project, part of The Illumos Foundation > > > > (C) > > > > 2010- > > > > 2022 > > > > Use is subject to license terms. > > > > Assembled 01 November 2022 > > > > > > > > Client is Linux Mint... > > > > Distributor ID: Linuxmint > > > > Description:Linux Mint 21 > > > > Release:21 > > > > Codename: vanessa > > > > > > > > Mount version is... > > > > mount from util-linux 2.37.2 (libmount 2.37.2: selinux, smack, > > > > btrfs, > > > > verity, namespaces, assert, debug) > > > > > > > > Mounting command is... > > > > sudo mount //192.168.0.2/jaguar /mnt/jaguar -o > > > > user=username,password=password,file_mode=0777,dir_mode=0777,ve > > > > rs=3 > > > > .0 > > > > > > > > Errors received in syslog are... > > > > > > > > Jul 21 12:34:04 main-desktop kernel: [1487413.251478] CIFS: > > > > __readahead_batch() returned 9/16 > > > > Jul 21 12:34:04 main-desktop kernel: [1487413.260788] CIFS: > > > > __readahead_batch() returned 13/16 > > > > Jul 21 12:34:04 main-desktop kernel: [1487413.668432] CIFS: > > > > __readahead_batch() returned 9/16 > > > > Jul 21 12:34:05 main-desktop kernel: [1487413.873347] CIFS: > > > > __readahead_batch() returned 3/16 > > > > Jul 21 12:34:05 main-desktop kernel: [1487413.877899] CIFS: > > > > __readahead_batch() returned 7/16 > > > > Jul 21 12:34:05 main-desktop kernel: [1487413.882006] CIFS: > > > > __readahead_batch() returned 9/16 > > > > Jul 21 12:34:05 main-desktop kernel: [1487413.887425] CIFS: > > > > __readahead_batch() returned 4/16 > > > > Jul 21 12:34:05 main-desktop kernel: [1487414.095730] CIFS: > > > > __readahead_batch() returned 12/16 > > > > Jul 21 12:34:05 main-desktop kernel: [1487414.303237] CIFS: > > > > __readahead_batch() returned 3/16 > > > > > > > > ___ > > > > openindiana-discuss mailing list > > > > openindiana-discuss@openindiana.org > > > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > > > > > ___ > > > openindiana-discuss mailing list > > > openindiana-discuss@openindiana.org > > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > --ms040004080303020505000105-- > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Readahead Batch error
Hi Till, Sorry ... to start from the beginning... When I buy a Blu Ray or DVD, I digitise them because many of my devices don't actually have drives these days. They are, "dumped," to a large network share on one of my OI servers. In order to sort them, I have to play them and then rename them once I've identified what they are. However, trying to rename them now that I've moved from 1.0 results in a "Resource Temporarily Unavailable" error message and I have to terminate the video playing, before I can rename. I looked up the error in syslog and assumed it was related... which it appears that it might do. The cache=none does appear to have solved the errors in the syslog, but my ability to rename a file when it's playing, is still happening. Confusingly I have two OI servers, built with the same version, at the same time, and they connect at different vers. One reports 3.0 and the other reports at 3.1.1 and the mount commands are the same, so I've got no clue as to how they are connecting at different versions. I've tried to look up mans for "mount" with the cache option, and I can't find any that explain that particular option. I think I'll go find a brick wall somewhere and bash my head against it for a while :-) Michelle. On Fri, 2023-07-21 at 19:37 +0200, Till Wegmüller wrote: > Hey Michelle > > The Internet tells me this might be related to linuxes client > caching. > Try cache=none if you are having locking problems. And experiment a > bit > with the cache variable. > > -Till > On 21.07.23 13:41, Michelle wrote: > > Can't work out why I'm getting this, which is causing problems > > accessing/renaming/moving files, more specifically if they're in > > use by > > another process, which used to be no problem under vers=1.0 > > > > ...but obviously I'm not going to shift back to vers=1.0. I did try > > connecting at 2.0 and also leaving the vers off the command > > (detailed > > later) but no change in behaviour. > > > > Looking the error up on the internet has not given anything that I > > can > > believe is relevant to this particular combination of client and > > server, etc. > > > > OI server is at... > > OpenIndiana Hipster 2022.10 (powered by illumos) > > OpenIndiana Project, part of The Illumos Foundation (C) > > 2010- > > 2022 > > Use is subject to license terms. > > Assembled 01 November 2022 > > > > Client is Linux Mint... > > Distributor ID: Linuxmint > > Description:Linux Mint 21 > > Release:21 > > Codename: vanessa > > > > Mount version is... > > mount from util-linux 2.37.2 (libmount 2.37.2: selinux, smack, > > btrfs, > > verity, namespaces, assert, debug) > > > > Mounting command is... > > sudo mount //192.168.0.2/jaguar /mnt/jaguar -o > > user=username,password=password,file_mode=0777,dir_mode=0777,vers=3 > > .0 > > > > Errors received in syslog are... > > > > Jul 21 12:34:04 main-desktop kernel: [1487413.251478] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:04 main-desktop kernel: [1487413.260788] CIFS: > > __readahead_batch() returned 13/16 > > Jul 21 12:34:04 main-desktop kernel: [1487413.668432] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.873347] CIFS: > > __readahead_batch() returned 3/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.877899] CIFS: > > __readahead_batch() returned 7/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.882006] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.887425] CIFS: > > __readahead_batch() returned 4/16 > > Jul 21 12:34:05 main-desktop kernel: [1487414.095730] CIFS: > > __readahead_batch() returned 12/16 > > Jul 21 12:34:05 main-desktop kernel: [1487414.303237] CIFS: > > __readahead_batch() returned 3/16 > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Readahead Batch error
Hi Till, Unfortunately it didn't work. I tested by running a video and then attempting to rename it while it was running. Previously it worked on ver 1.0 but even with cache=none and ver=3.0 it failed to rename with the resource temporarily unavailable. I'll read up on the cache variable and see what options I can try. Michelle. On Fri, 2023-07-21 at 19:37 +0200, Till Wegmüller wrote: > Hey Michelle > > The Internet tells me this might be related to linuxes client > caching. > Try cache=none if you are having locking problems. And experiment a > bit > with the cache variable. > > -Till > On 21.07.23 13:41, Michelle wrote: > > Can't work out why I'm getting this, which is causing problems > > accessing/renaming/moving files, more specifically if they're in > > use by > > another process, which used to be no problem under vers=1.0 > > > > ...but obviously I'm not going to shift back to vers=1.0. I did try > > connecting at 2.0 and also leaving the vers off the command > > (detailed > > later) but no change in behaviour. > > > > Looking the error up on the internet has not given anything that I > > can > > believe is relevant to this particular combination of client and > > server, etc. > > > > OI server is at... > > OpenIndiana Hipster 2022.10 (powered by illumos) > > OpenIndiana Project, part of The Illumos Foundation (C) > > 2010- > > 2022 > > Use is subject to license terms. > > Assembled 01 November 2022 > > > > Client is Linux Mint... > > Distributor ID: Linuxmint > > Description:Linux Mint 21 > > Release:21 > > Codename: vanessa > > > > Mount version is... > > mount from util-linux 2.37.2 (libmount 2.37.2: selinux, smack, > > btrfs, > > verity, namespaces, assert, debug) > > > > Mounting command is... > > sudo mount //192.168.0.2/jaguar /mnt/jaguar -o > > user=username,password=password,file_mode=0777,dir_mode=0777,vers=3 > > .0 > > > > Errors received in syslog are... > > > > Jul 21 12:34:04 main-desktop kernel: [1487413.251478] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:04 main-desktop kernel: [1487413.260788] CIFS: > > __readahead_batch() returned 13/16 > > Jul 21 12:34:04 main-desktop kernel: [1487413.668432] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.873347] CIFS: > > __readahead_batch() returned 3/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.877899] CIFS: > > __readahead_batch() returned 7/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.882006] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.887425] CIFS: > > __readahead_batch() returned 4/16 > > Jul 21 12:34:05 main-desktop kernel: [1487414.095730] CIFS: > > __readahead_batch() returned 12/16 > > Jul 21 12:34:05 main-desktop kernel: [1487414.303237] CIFS: > > __readahead_batch() returned 3/16 > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Readahead Batch error
In case it helps, this is what's reported by mount... //192.168.0.2/jaguar on /mnt/jaguar type cifs (rw,relatime,vers=3.0,cache=none,username=michelle,uid=0,noforceuid,gid =0,noforcegid,addr=192.168.0.2,file_mode=0777,dir_mode=0777,soft,nounix ,serverino,mapposix,rsize=65536,wsize=65536,bsize=1048576,echo_interval =60,actimeo=1,user=michelle) Michelle. On Fri, 2023-07-21 at 19:37 +0200, Till Wegmüller wrote: > Hey Michelle > > The Internet tells me this might be related to linuxes client > caching. > Try cache=none if you are having locking problems. And experiment a > bit > with the cache variable. > > -Till > On 21.07.23 13:41, Michelle wrote: > > Can't work out why I'm getting this, which is causing problems > > accessing/renaming/moving files, more specifically if they're in > > use by > > another process, which used to be no problem under vers=1.0 > > > > ...but obviously I'm not going to shift back to vers=1.0. I did try > > connecting at 2.0 and also leaving the vers off the command > > (detailed > > later) but no change in behaviour. > > > > Looking the error up on the internet has not given anything that I > > can > > believe is relevant to this particular combination of client and > > server, etc. > > > > OI server is at... > > OpenIndiana Hipster 2022.10 (powered by illumos) > > OpenIndiana Project, part of The Illumos Foundation (C) > > 2010- > > 2022 > > Use is subject to license terms. > > Assembled 01 November 2022 > > > > Client is Linux Mint... > > Distributor ID: Linuxmint > > Description:Linux Mint 21 > > Release:21 > > Codename: vanessa > > > > Mount version is... > > mount from util-linux 2.37.2 (libmount 2.37.2: selinux, smack, > > btrfs, > > verity, namespaces, assert, debug) > > > > Mounting command is... > > sudo mount //192.168.0.2/jaguar /mnt/jaguar -o > > user=username,password=password,file_mode=0777,dir_mode=0777,vers=3 > > .0 > > > > Errors received in syslog are... > > > > Jul 21 12:34:04 main-desktop kernel: [1487413.251478] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:04 main-desktop kernel: [1487413.260788] CIFS: > > __readahead_batch() returned 13/16 > > Jul 21 12:34:04 main-desktop kernel: [1487413.668432] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.873347] CIFS: > > __readahead_batch() returned 3/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.877899] CIFS: > > __readahead_batch() returned 7/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.882006] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.887425] CIFS: > > __readahead_batch() returned 4/16 > > Jul 21 12:34:05 main-desktop kernel: [1487414.095730] CIFS: > > __readahead_batch() returned 12/16 > > Jul 21 12:34:05 main-desktop kernel: [1487414.303237] CIFS: > > __readahead_batch() returned 3/16 > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Readahead Batch error
Hi Till, Thanks very much. I'll give that a try. Michelle. On Fri, 2023-07-21 at 19:37 +0200, Till Wegmüller wrote: > Hey Michelle > > The Internet tells me this might be related to linuxes client > caching. > Try cache=none if you are having locking problems. And experiment a > bit > with the cache variable. > > -Till > On 21.07.23 13:41, Michelle wrote: > > Can't work out why I'm getting this, which is causing problems > > accessing/renaming/moving files, more specifically if they're in > > use by > > another process, which used to be no problem under vers=1.0 > > > > ...but obviously I'm not going to shift back to vers=1.0. I did try > > connecting at 2.0 and also leaving the vers off the command > > (detailed > > later) but no change in behaviour. > > > > Looking the error up on the internet has not given anything that I > > can > > believe is relevant to this particular combination of client and > > server, etc. > > > > OI server is at... > > OpenIndiana Hipster 2022.10 (powered by illumos) > > OpenIndiana Project, part of The Illumos Foundation (C) > > 2010- > > 2022 > > Use is subject to license terms. > > Assembled 01 November 2022 > > > > Client is Linux Mint... > > Distributor ID: Linuxmint > > Description:Linux Mint 21 > > Release:21 > > Codename: vanessa > > > > Mount version is... > > mount from util-linux 2.37.2 (libmount 2.37.2: selinux, smack, > > btrfs, > > verity, namespaces, assert, debug) > > > > Mounting command is... > > sudo mount //192.168.0.2/jaguar /mnt/jaguar -o > > user=username,password=password,file_mode=0777,dir_mode=0777,vers=3 > > .0 > > > > Errors received in syslog are... > > > > Jul 21 12:34:04 main-desktop kernel: [1487413.251478] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:04 main-desktop kernel: [1487413.260788] CIFS: > > __readahead_batch() returned 13/16 > > Jul 21 12:34:04 main-desktop kernel: [1487413.668432] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.873347] CIFS: > > __readahead_batch() returned 3/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.877899] CIFS: > > __readahead_batch() returned 7/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.882006] CIFS: > > __readahead_batch() returned 9/16 > > Jul 21 12:34:05 main-desktop kernel: [1487413.887425] CIFS: > > __readahead_batch() returned 4/16 > > Jul 21 12:34:05 main-desktop kernel: [1487414.095730] CIFS: > > __readahead_batch() returned 12/16 > > Jul 21 12:34:05 main-desktop kernel: [1487414.303237] CIFS: > > __readahead_batch() returned 3/16 > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Readahead Batch error
Can't work out why I'm getting this, which is causing problems accessing/renaming/moving files, more specifically if they're in use by another process, which used to be no problem under vers=1.0 ...but obviously I'm not going to shift back to vers=1.0. I did try connecting at 2.0 and also leaving the vers off the command (detailed later) but no change in behaviour. Looking the error up on the internet has not given anything that I can believe is relevant to this particular combination of client and server, etc. OI server is at... OpenIndiana Hipster 2022.10 (powered by illumos) OpenIndiana Project, part of The Illumos Foundation (C) 2010- 2022 Use is subject to license terms. Assembled 01 November 2022 Client is Linux Mint... Distributor ID: Linuxmint Description:Linux Mint 21 Release:21 Codename: vanessa Mount version is... mount from util-linux 2.37.2 (libmount 2.37.2: selinux, smack, btrfs, verity, namespaces, assert, debug) Mounting command is... sudo mount //192.168.0.2/jaguar /mnt/jaguar -o user=username,password=password,file_mode=0777,dir_mode=0777,vers=3.0 Errors received in syslog are... Jul 21 12:34:04 main-desktop kernel: [1487413.251478] CIFS: __readahead_batch() returned 9/16 Jul 21 12:34:04 main-desktop kernel: [1487413.260788] CIFS: __readahead_batch() returned 13/16 Jul 21 12:34:04 main-desktop kernel: [1487413.668432] CIFS: __readahead_batch() returned 9/16 Jul 21 12:34:05 main-desktop kernel: [1487413.873347] CIFS: __readahead_batch() returned 3/16 Jul 21 12:34:05 main-desktop kernel: [1487413.877899] CIFS: __readahead_batch() returned 7/16 Jul 21 12:34:05 main-desktop kernel: [1487413.882006] CIFS: __readahead_batch() returned 9/16 Jul 21 12:34:05 main-desktop kernel: [1487413.887425] CIFS: __readahead_batch() returned 4/16 Jul 21 12:34:05 main-desktop kernel: [1487414.095730] CIFS: __readahead_batch() returned 12/16 Jul 21 12:34:05 main-desktop kernel: [1487414.303237] CIFS: __readahead_batch() returned 3/16 ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] USB Protocol
I had an incident with a backup set and I thought I'd ask if there are any extra steps/commands I can take/do when adding and removing USB external drives. Basically, HP Gen 10 Plus running the latest Hipster, updated a few weeks ago. Hooking up an 8th external drive with a ZFS set on it (I actually have two) ... and I had previously backed up to both of them with rsync a week-ish ago. Come last night, I hook up one drive, import the zpool, do the rsync, export the rpool and switch off the caddy. Then put the other drive in, switch the caddy on and attempt an import, but there was no pool on the drive. I had to re-create the pool and do a full rsync. I'm putting this down to bad hygine on my part and I was wondering whether there is anything I can do to improve my handling of using external USB caddys for backup. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Best way to install rsync
No worries... pkg install rsync It was already in the library! I should have tried that first. Michelle. On Fri, 2023-03-31 at 15:43 +0100, Michelle wrote: > I've got OI minimal install loaded and rsync isn't there. > > Some sources are suggesting downloading from Oracle. Others to > compile > it. > > I was wondering what would people recommend in order to get rsync > onto > the installation please? > > Michelle. > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Best way to install rsync
I've got OI minimal install loaded and rsync isn't there. Some sources are suggesting downloading from Oracle. Others to compile it. I was wondering what would people recommend in order to get rsync onto the installation please? Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Default SMB file permission
Thanks very much for taking the time to go into all this detail. I'll grab a cuppa and give it a thorough read. Michelle. On Fri, 2023-03-31 at 02:38 +0200, Guenther Alka wrote: > Main open question is if you use SAMBA or the Solaris kernelbased > SMB > server. > I would always prefer the second due easier config and better > handling > of Windows SMB permissions > and zero config ZFS snaps= Windows previous versions. > > Due the lack of smb.conf, I asume you use the kernelbased SMB server: > - Settings are done via ZFS properties aclmode and aclinherit or > sharectl > https://illumos.org/man/8/sharectl > > or in napp-it > > > The kernelbased SMB server use (only and always) Windows ntfs alike > permissions > with inheritance based on Windows SID as security reference. This is > why > permissions ex > in an AD environment remaion intact after a restore from backup > without > any mappings. > https://forums.servethehome.com/index.php?threads/napp-it-zfs-server-on-omnios-solaris-news-tips-and-tricks.38240/#post-368594 > > Additionally to Unix groups you have Windows alike local SMB groups. > > ACL are superiour to classic Unix permissions like 750. > Never set classic permissions or gid as they delete inheritance > settings or reduce permissions, always use ACL. Permissions wise > Solaris > is like Windows not Unix. > > To set ACL > - set aclinherit to pass-through (Windows alike) > - prefer Windows. SMB connect as root and set ACL > > - for the shared filesystem: > allow at least read access for the shared folder only > > for folders below > set needed settings with inheritance to files and folders > ex modify for certain users/ groups > > A possible default is also: > - allow read for everyone@ (shared filesystem, this folder only) > - allow creation of files and folders for everyone > Default is then that a creator (=owner) has full permissions, others > lack permissions > > Additionally you can set share ACL. When you (re)create a share, the > are > always everyone@=full > Napp-it can store/ recreate share permissions as private ZFS > properties > on a re-share > https://forums.servethehome.com/index.php?threads/napp-it-zfs-server-on-omnios-solaris-news-tips-and-tricks.38240/#post-356373 > > Gea > > > > more, > https://forums.servethehome.com/index.php?threads/napp-it-zfs-server-on-omnios-solaris-news-tips-and-tricks.38240/#post-356845 > > I'm bashing my head a little. > > > > I have a newly built OI server and I've got a share out through ZFS > > smb > > share which is being accessed from a Linux client. > > > > Every file written comes in with 700 and I need to change that > > default > > to 740. > > > > I believed that was a setting in smb.conf rather than umask but I > > think > > I'm getting my linux and unix mixed up. But I can't find smb.conf > > anyway. > > > > Oddly the file does show -rwx--+ which potentially indicates an > > acl > > applying > > > > The guid bit for directory permission inheritance is working. It's > > just > > newly created files. > > > > I'm going to lie down because my head is hurting. > > > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Default SMB file permission
I'm bashing my head a little. I have a newly built OI server and I've got a share out through ZFS smb share which is being accessed from a Linux client. Every file written comes in with 700 and I need to change that default to 740. I believed that was a setting in smb.conf rather than umask but I think I'm getting my linux and unix mixed up. But I can't find smb.conf anyway. Oddly the file does show -rwx--+ which potentially indicates an acl applying The guid bit for directory permission inheritance is working. It's just newly created files. I'm going to lie down because my head is hurting. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] VIM and HP cert
The certificate part of the question has been solved. Apparently when clicking "remove certificate" it pops up further notice that it will not only remove the certificate, but it will generate a new one automatically and reset iLO. Bingo. Another 5 years certificate. Michelle. On Wed, 2023-03-29 at 22:12 +0100, Michelle wrote: > I just bought myself a new (to me) HP Gen 10 Plus and put OI-Hipster- > minimal-20221123 on it. > > On using crontab -e I get... > E1187: Failed to source defaults.vim > > A lot of conversations on the web say this is a bug in a certain > build > and that it was fixed last year. I was wondering if it's just got to > make its way into OI at some point, or perhaps I should just update > the > system? Or possibly if another editor is installed that would be > easier > than VI commands :-) > > Thought I'd ask people's thoughts before I do anything. > > Also, this server has iLO 5 but the default self cert on the server > only lasts five years. While I've used certs with proper trust > chains, > the one for this would need to be self-signed, but I'm presuming I'm > going to have to put some form of cert trust in first... I'm checking > with people first because if I get this wrong I could seriously bork > iLO and stuff myself up something chronic. > > ... or is there a way to SSH to iLO and get it to generate its own > certificate for 20 years or something. > > Self signed certs can be a whole world of pain if I get it wrong. > > Thanks in advance for any thoughts and advice. > > Michelle. > > > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] VIM and HP cert
I just bought myself a new (to me) HP Gen 10 Plus and put OI-Hipster- minimal-20221123 on it. On using crontab -e I get... E1187: Failed to source defaults.vim A lot of conversations on the web say this is a bug in a certain build and that it was fixed last year. I was wondering if it's just got to make its way into OI at some point, or perhaps I should just update the system? Or possibly if another editor is installed that would be easier than VI commands :-) Thought I'd ask people's thoughts before I do anything. Also, this server has iLO 5 but the default self cert on the server only lasts five years. While I've used certs with proper trust chains, the one for this would need to be self-signed, but I'm presuming I'm going to have to put some form of cert trust in first... I'm checking with people first because if I get this wrong I could seriously bork iLO and stuff myself up something chronic. ... or is there a way to SSH to iLO and get it to generate its own certificate for 20 years or something. Self signed certs can be a whole world of pain if I get it wrong. Thanks in advance for any thoughts and advice. Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] OI on HP Gen 10 Microserver
Hi Rolf, Thanks for this. What I find interesting is that looking at the data sheet, etc. there is no clear indication that two are 6G and 2 are 3G. They aren't exactly up front about it. That's what's worrying me about the 10+, I can't find anything that spells it out what the ports are capable of. However, it does appear that the 10 is a Marvell controller and doesn't advertise raid 5, whereas the 10+ is a Smart Array S100i controller and supports raid 5, so I would have believed that they will all be 6G. If they advertised RAID 5 and one of the drives was operating at half the speed, then they'd likely be in for a bit of a backlash I would have thought. The only way to find out for sure... if someone doesn't already know... is to wait until I've got it, put some drives in and see what the BIOS says that the drives are connected at. At the moment, given the bug that Carsten mentioned with booting rpool from USB, I'm thinking of using one bay for SSD (with an SFF LFF adaptor caddy) and the other three for the spinning rust. Backup drives will just have to be attached by USB 3 caddy, I guess. Michelle. On Thu, 2023-02-16 at 22:21 +0100, Rolf M. Dietze wrote: > Hi Michelle, > > don't know about the G10+ boxes. I meant: > https://buy.hpe.com/uk/en/servers/proliant-microserver/proliant-microserver/proliant-microserver/hpe-proliant-microserver-gen10/p/1009955118 > and with that issue of 2 ports 6G and 2 ports 3G I rather looked into > boxes from other manufacturers than HP to recomend them to customers. > > With the older Gen10 boxes one needs .NET for a console via the build > in system controler, even if you purchase the advanced license. Since > there is no windows in the net at all, I set them up on a KVM server > and never bought any boxes fom HP because of this rather flimsy > design > (no console via systemcontroller, 6G/3G sata ports) > > But if I have to choose between OI and FreeBSD and what experiences > running OI as a server OS on those boxes, I go for OI and that was > the original discussion > > regs, Rolf > > > Quoting Michelle : > > > The target server is actually going to be the Gen 10 Plus... > > https://uk-new.ingrammicro.com/site/productdetail?id=2M2SBG9 > > > > The specs are light on the exact specs of the speed, but the manual > > doesn't help either... > > https://www.serversdirect.co.uk/PDFs/P16006-421_1_7006250_Original.pdf?v=27 > > > > The best they say is... "The embedded HPE Smart Array S100i SR > > Gen10 > > Controller supports SATA drive installation. This controller > > supports > > RAID levels 0, 1, 5, and 10" ... and it would be a bit odd for them > > to > > support RAID levels like that and have two ports down to 3Gbs. > > > > Grateful for any insight. > > > > Michelle. > > > > > > On Thu, 2023-02-16 at 19:55 +0100, rmd wrote: > > > Hi > > > > > > it depends a bit on the BIOS version of the Gen 10. With the > > > early > > > version > > > it works best, pulling all disks out of the bays, do the install, > > > reboot, halt > > > again, put the drives back in, repower and boot, then do the disk > > > configs. > > > > > > With my Gen 10, I put a 4i4o SAS-controler in it, plug the sas- > > > plug > > > of the > > > internal drive bays on the i-cnector of that additional 4i4o sas > > > card > > > and > > > connect SSDs with a fan out cable to the on board sas connector. > > > With > > > a bit of soldering an additional power connector for the SSDs is > > > simply > > > set up an the OS goes on those SSDs. Hence, the Gen10 supports > > > only > > > 2 diskbays with 6G througput and the other two bays only support > > > 3G. > > > > > > regs, Rolf > > > > > > > On Feb 16, 2023, at 9:34 AM, Michelle > > > > wrote: > > > > > > > > One of my older servers has died and I'm looking to replace > > > > with a > > > > Gen > > > > 10 microserver. > > > > > > > > On my N54L, there are four bays, but a top CD bay where the > > > > operating > > > > system drive sits with OI on it, leaving me with four bays for > > > > spinning > > > > rust. > > > > > > > > However, on the Gen 10 I've looked at the motherboard and I > > > > can't > > > > see > > > > anything other than a USB connector and the SATA h
Re: [OpenIndiana-discuss] OI on HP Gen 10 Microserver
The target server is actually going to be the Gen 10 Plus... https://uk-new.ingrammicro.com/site/productdetail?id=2M2SBG9 The specs are light on the exact specs of the speed, but the manual doesn't help either... https://www.serversdirect.co.uk/PDFs/P16006-421_1_7006250_Original.pdf?v=27 The best they say is... "The embedded HPE Smart Array S100i SR Gen10 Controller supports SATA drive installation. This controller supports RAID levels 0, 1, 5, and 10" ... and it would be a bit odd for them to support RAID levels like that and have two ports down to 3Gbs. Grateful for any insight. Michelle. On Thu, 2023-02-16 at 19:55 +0100, rmd wrote: > Hi > > it depends a bit on the BIOS version of the Gen 10. With the early > version > it works best, pulling all disks out of the bays, do the install, > reboot, halt > again, put the drives back in, repower and boot, then do the disk > configs. > > With my Gen 10, I put a 4i4o SAS-controler in it, plug the sas-plug > of the > internal drive bays on the i-cnector of that additional 4i4o sas card > and > connect SSDs with a fan out cable to the on board sas connector. With > a bit of soldering an additional power connector for the SSDs is > simply > set up an the OS goes on those SSDs. Hence, the Gen10 supports only > 2 diskbays with 6G througput and the other two bays only support 3G. > > regs, Rolf > > > On Feb 16, 2023, at 9:34 AM, Michelle > > wrote: > > > > One of my older servers has died and I'm looking to replace with a > > Gen > > 10 microserver. > > > > On my N54L, there are four bays, but a top CD bay where the > > operating > > system drive sits with OI on it, leaving me with four bays for > > spinning > > rust. > > > > However, on the Gen 10 I've looked at the motherboard and I can't > > see > > anything other than a USB connector and the SATA harness connector. > > I > > can't see a DVD/CD drive either. > > > > Has anyone managed to get OI on another drive, or am I going to > > have to > > take up a bay for the OS drive please? > > > > Thanks for any hints and advice, > > > > Michelle. > > > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] OI on HP Gen 10 Microserver
Hi Russell, That's great. Cheers for that. Is there any advice on running a server for years from a USB stick please? I'm thinking of an SSD hard drive on a USB adaptor instead. Not sure yet. Grateful for any thoughts and information that could help me make up my mind. Michelle. On Thu, 2023-02-16 at 13:13 +, russell wrote: > Hi Michelle > > If you go here you can download a Live USB image to boot from > https://www.openindiana.org/download/ > > Kind Regards > > Russell > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] OI on HP Gen 10 Microserver
One of my older servers has died and I'm looking to replace with a Gen 10 microserver. On my N54L, there are four bays, but a top CD bay where the operating system drive sits with OI on it, leaving me with four bays for spinning rust. However, on the Gen 10 I've looked at the motherboard and I can't see anything other than a USB connector and the SATA harness connector. I can't see a DVD/CD drive either. Has anyone managed to get OI on another drive, or am I going to have to take up a bay for the OS drive please? Thanks for any hints and advice, Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS tolerance
Hi Till, Thanks for the feedback. I'm deliberately trying to push the envelope. I believe that ZFS will act as a buffer between drives and run at the slowest device speed, but I'm wondering if there is a threshold where ever ZFS will throw up its hands and give up. I'm trying to make a ZFS pool focused on resilience, while using drives that are already laying around. Performance is understandably going to take a hit. The vision is a pool consisting of a number of mirrored vdevs. I mean... if I had a load of hard drives on SATA and connected another drive via SATA to USB adaptor, would ZFS still handle it and slow down, or would it throw a fit? Many thanks, Michelle. On Sat, 2022-10-29 at 12:33 -0300, Till Wegmueller wrote: > Hi > > It can get yarring. Try to make raids with drives that are similar > and > combine them into a mutt pool then you should get the least effect. > Otherwise the slowest drive may dictate the pools performance. > > There are more tuning options, but I hope this helps a little. > Greetings > Till > > On 29/10/2022 10.36, Michelle wrote: > > Hi All, > > > > I am doing a small idiotic project, putting together odds and sods > > of > > hard drives to make a RAID set. You know the kind of thing, the > > hard > > drives that are remnants of old projects but you don't want to > > throw > > them out because... well... they work! > > > > So I was wondering that in a pile of hard drives, 5400 to 7200 RPM, > > green and red drives than spin down and up again, and blue/black > > drives > > that keep running > > > > Just how much tolerance does ZFS have regarding the timings of > > using > > different devices? Particularly when under sustained transfer to > > the > > server. > > > > My target will be RAID 0+1 with partitions on larger drives, to > > pair > > against smaller drives. > > > > Grateful for any thoughts/information please. > > > > Many thanks, > > > > Michelle. > > > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] ZFS tolerance
Hi All, I am doing a small idiotic project, putting together odds and sods of hard drives to make a RAID set. You know the kind of thing, the hard drives that are remnants of old projects but you don't want to throw them out because... well... they work! So I was wondering that in a pile of hard drives, 5400 to 7200 RPM, green and red drives than spin down and up again, and blue/black drives that keep running Just how much tolerance does ZFS have regarding the timings of using different devices? Particularly when under sustained transfer to the server. My target will be RAID 0+1 with partitions on larger drives, to pair against smaller drives. Grateful for any thoughts/information please. Many thanks, Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS 0+1 across disparate drives
It is a possibility, but LTO-8 drives are still £3,000+ for just the drive. The drive would also then be the weakest link, and quite an expensive one. With 10Tb NAS drives coming in at £300 each, that's £1,200 for four drives with 30Tb capacity at RaidZ. An extra as spare, or two extras for RaidZ2 and a spare, and you still haven't broken £2,000. Michelle. On Thu, 2021-12-02 at 19:55 +0100, s...@pandora.be wrote: > I have only read this thread very superficially, but LTO tape (linear > tape open) > > https://en.wikipedia.org/wiki/Linear_Tape-Open > > LTO-8 has native raw capacity of about 12TB so attaching a LTO-8 > drive may be an option. > > The same page says, by the way, that LTO-9 is supposed to have 18TB > native capacity, > but "delayed to late 2021" (which is now). > > Given the fact that the compressed capacity of LTO-8 is 30TB you are > probably still well below the limits of a single LTO8 tape, > that is, it is quite possible that if you have less than 30TB of > data, you can make a backup to a LTO-8 drive. > > Regards, > David Stes > > - Op 1 dec 2021 om 18:08 schreef Mehmet Erol Sanliturk > m.e.sanlit...@gmail.com: > > > On Wed, Dec 1, 2021 at 2:55 PM Michelle > > wrote: > > > > > I'm trying to achieve a resilient way of bringing together all my > > > older > > > drives for a backup solution using scraps of whatever I can get > > > my > > > hands on. > > > > > > I have closing on 12TB of data so even the 10 won't be enough to > > > back > > > everything up, but this is as much for the exercise of doing it, > > > as > > > achieving anything solid. It won't be under pressure, but I'd > > > rather > > > push the envelope and see what I can do. > > > > > > So how would the command go? > > > > > > zpool create tank raidz mirror drive1 drive2 mirror drive3 drive4 > > > drive5 > > > > > > ...which is where I come unstuck with the 2TB drive in the mix. > > > > > > > > > > > > > > > > > A few days ago I lost many weeks of work because my drive #1 / 3 > > died , not > > synchronized into #2 / 3 and #3 / 3 . > > This made the computer unbootable . > > I replaced the failed disk and synchronized it with #2 . > > The disk #3 failed and made the computer unbootable . > > I replaced that disk and synchronized it with #2 . > > > > The #1 disk was new but bought approximately 5 year ago . > > > > You are saying your disks are older . > > > > One "safe" but slow choice would be the following . > > > > Use external USB docks for each of your disks and make their file > > systems > > compatible > > with your computer ( if your disks have other file systems ) . > > > > With your synchronization shell scripts (1) mount (2) rsync (3) > > un_mount > > your drives > > by using cron or ( manually which this option is not a good choice > > ) . > > > > If any one of your disks fails , it will not affect your computer . > > This will be slow but without any other harm . > > > > > > The most suitable additional action may be to backup your data to > > external > > disks > > regularly . These disks will not be continuously connected to the > > computer > > and will not be affected by electricity harmful effects . > > > > > > > > OR > > > > You may use another computer ( such as a single board computer ) as > > an NFS > > server > > ( or a NAS if one is available to you ) , and use your drives in > > that > > server . > > Then synchronize your drives from your computer . If any disk fails > > , it > > only affects the server > > but your computer continues to work without affection . > > > > > > > > Mehmet Erol Sanliturk > > > > > > > > > > > > > > > > > > > > > > > > > On Wed, 2021-12-01 at 11:17 +, James wrote: > > > > On 01/12/2021 08:31, Michelle wrote: > > > > > Say I was to put a 2tb, three 4tb and a 6tb together (a 2 and > > > > > two 4 > > > > > would make 10 and the other 4 and the 6 would also make 10) > > > > > > > > > > Would that be possible with ZFS now? > > > > > > > > I think it has always been possible, ask is is sensible? Try > > > > it, if > > > > you > &g
Re: [OpenIndiana-discuss] ZFS 0+1 across disparate drives
Good advice. Thanks for taking the time. This remains one question ... can drives and partitions be mixed in a ZFS pool? This project is a backup project which is actually a backup of a backup, if that makes sense. The original data is spread across two servers, one running OI (the main data store) and a Cisco NAS (my video/rip store). These are backed up to external drives, with two backup sets for each. So I already have a situation with one RAID resilient copy of the data, along with two full external backups of the data. But I have a series of odd drives hanging around, as many people probably do. And I want to do something useful with them. The only other alternative is to dismantle and trash the drives. So I am going to undertaken this project, and document it, so that others in the same situation can find some use for their drives as opposed to simply throwing away perfectly serviceable drives. Yes, the drives are going to be disparate. Most of them are WD Red units, one is WD green. For this project I will also be re-using an old motherboard which has been sitting in the attic for years. The operating system will be on its own SSD drive, and the whole thing will be documented so in the case of an OS failure, the drive can easily be rebuilt and reconfigured... because all the notes are there. So... yes... partially I'm doing this for the hell of it. And hopefully my success or failure will be documented and of use to other people. I have a small youtube channel - not monetized - where all this will ultimately be recorded. On Wed, 2021-12-01 at 20:08 +0300, Mehmet Erol Sanliturk wrote: > On Wed, Dec 1, 2021 at 2:55 PM Michelle > wrote: > > > I'm trying to achieve a resilient way of bringing together all my > > older > > drives for a backup solution using scraps of whatever I can get my > > hands on. > > > > I have closing on 12TB of data so even the 10 won't be enough to > > back > > everything up, but this is as much for the exercise of doing it, as > > achieving anything solid. It won't be under pressure, but I'd > > rather > > push the envelope and see what I can do. > > > > So how would the command go? > > > > zpool create tank raidz mirror drive1 drive2 mirror drive3 drive4 > > drive5 > > > > ...which is where I come unstuck with the 2TB drive in the mix. > > > > > > > > > > A few days ago I lost many weeks of work because my drive #1 / 3 died > , not > synchronized into #2 / 3 and #3 / 3 . > This made the computer unbootable . > I replaced the failed disk and synchronized it with #2 . > The disk #3 failed and made the computer unbootable . > I replaced that disk and synchronized it with #2 . > > The #1 disk was new but bought approximately 5 year ago . > > You are saying your disks are older . > > One "safe" but slow choice would be the following . > > Use external USB docks for each of your disks and make their file > systems > compatible > with your computer ( if your disks have other file systems ) . > > With your synchronization shell scripts (1) mount (2) rsync (3) > un_mount > your drives > by using cron or ( manually which this option is not a good choice ) > . > > If any one of your disks fails , it will not affect your computer . > This will be slow but without any other harm . > > > The most suitable additional action may be to backup your data to > external > disks > regularly . These disks will not be continuously connected to the > computer > and will not be affected by electricity harmful effects . > > > > OR > > You may use another computer ( such as a single board computer ) as > an NFS > server > ( or a NAS if one is available to you ) , and use your drives in that > server . > Then synchronize your drives from your computer . If any disk fails , > it > only affects the server > but your computer continues to work without affection . > > > > Mehmet Erol Sanliturk > > > > > > > > > > > > > On Wed, 2021-12-01 at 11:17 +, James wrote: > > > On 01/12/2021 08:31, Michelle wrote: > > > > Say I was to put a 2tb, three 4tb and a 6tb together (a 2 and > > > > two 4 > > > > would make 10 and the other 4 and the 6 would also make 10) > > > > > > > > Would that be possible with ZFS now? > > > > > > I think it has always been possible, ask is is sensible? Try it, > > > if > > > you > > > have nothing to loose. The problem is if one drive fails is > > > takes > > > out > > > all of one side of the mirror. &g
Re: [OpenIndiana-discuss] ZFS 0+1 across disparate drives
You're perfectly correct. After I put the command together, I realised that I'd made a mistake in putting raidz in there. The balance of partitions and using those instead of drives... I hadn't thought of that. I guess I'm just too used to thinking at the drive level for ZFS. On Wed, 2021-12-01 at 15:33 +, James wrote: > On 01/12/2021 11:55, Michelle wrote: > > I have closing on 12TB of data so even the 10 won't be enough to > > back > > everything up, but this is as much for the exercise of doing it, as > > achieving anything solid. It won't be under pressure, but I'd > > rather > > push the envelope and see what I can do. > > > > So how would the command go? > > > > zpool create tank raidz mirror drive1 drive2 mirror drive3 drive4 > > drive5 > > > > ...which is where I come unstuck with the 2TB drive in the mix. > > You are right to be stuck as ZFS can not make a mirror of raid0 in > one > command, zfs is wanting raid0 of mirrors which you can't get to 10TB > because you don't have pairs. You can pair the 4&6 and raid0 with > 4&4 > and get 8TB of continuous pool. > > Partition the 6 into 4 and 2, then you have pairs for raid0 of > mirrors. > Loose the 6TB and you loose 1/2 of 2 mirrors but you are better off > than > loosing 1/2 the entire mirror with mirror of raid0. 10TB using all > devices. > > Try making 2 raid0 pools 2+4+4 and 4+6, create volumes on those > pools > and use to make a mirror, /dev/zvol/dsk/etc... Untested and I can't > see > it being better. > > Use raidz1 of 4, 4, 4, and 6 = 12TB. This ignore 2TB of the 6 and > you > get a bonus doorstop by not using the 2TB. 3+1 is not a preferred > arrangement. > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS 0+1 across disparate drives
I'm trying to achieve a resilient way of bringing together all my older drives for a backup solution using scraps of whatever I can get my hands on. I have closing on 12TB of data so even the 10 won't be enough to back everything up, but this is as much for the exercise of doing it, as achieving anything solid. It won't be under pressure, but I'd rather push the envelope and see what I can do. So how would the command go? zpool create tank raidz mirror drive1 drive2 mirror drive3 drive4 drive5 ...which is where I come unstuck with the 2TB drive in the mix. On Wed, 2021-12-01 at 11:17 +, James wrote: > On 01/12/2021 08:31, Michelle wrote: > > Say I was to put a 2tb, three 4tb and a 6tb together (a 2 and two 4 > > would make 10 and the other 4 and the 6 would also make 10) > > > > Would that be possible with ZFS now? > > I think it has always been possible, ask is is sensible? Try it, if > you > have nothing to loose. The problem is if one drive fails is takes > out > all of one side of the mirror. > > Why not use 2 separate 4TB mirrors? 4&4 = 4, 4&6 = 4, total 8. You > loose the 2TB drive completely but (guessing) it is the slowest and > oldest. You ignore 2TB of the 6. > > You don't say what you are trying to achieve but it's unlikely you > have > the full 8TB of data, it is unlikely it can't be split. > > > > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS 0+1 across disparate drives
That's a workable solution. As you say, not as elegant as RAID, and if one drive went down on the live set, it would render the share inaccessible... but it would work in principle. Thanks for the suggestion. On Wed, 2021-12-01 at 13:22 +0300, Mehmet Erol Sanliturk wrote: > On Wed, Dec 1, 2021 at 11:32 AM Michelle > wrote: > > > This is probably going to be a stupid question, as I know that > > several > > year ago, the answer to this question would be no... but times > > change. > > > > Like many people, I have a collection of disparate drives sitting > > around doing nothing. > > > > I was thinking of getting them together in a RAID 1+0 > > configuration. > > > > Say I was to put a 2tb, three 4tb and a 6tb together (a 2 and two 4 > > would make 10 and the other 4 and the 6 would also make 10) > > > > Would that be possible with ZFS now? > > > > Michelle. > > > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > > Another possibility may be , perhaps , as follows : > > Attach these disks to your system(s) . > Prepare rsync containing shell script(s) . > Using cron , execute these shell script(s) to synchronize your disks > suitable to your needs. > Updates may not be instantaneous as RAID , but at least are working > with > respect to your design . > > > With my best wishes for all of you , > > > Mehmet Erol Sanliturk > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] ZFS 0+1 across disparate drives
This is probably going to be a stupid question, as I know that several year ago, the answer to this question would be no... but times change. Like many people, I have a collection of disparate drives sitting around doing nothing. I was thinking of getting them together in a RAID 1+0 configuration. Say I was to put a 2tb, three 4tb and a 6tb together (a 2 and two 4 would make 10 and the other 4 and the 6 would also make 10) Would that be possible with ZFS now? Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Rpool uncorrectable error
AHCI was already enabled. The IDE mode was legacy, so I've turned that off and a few other settings, but there's only so far I can go. So far, the OS is booting without the need for re-installation as it was already AHCI... but... well Michelle. On Sun, 2021-08-29 at 14:56 +0300, Toomas Soome via openindiana-discuss wrote: > > On 29. Aug 2021, at 14:31, Michelle wrote: > > > > I'm sat here, not knowing quite what I'm dealing with. > > > > OI recent build on HP N54L, as you may remember I hit a problem on > > the > > 8th August with rpool encountering an error on the SSD with the OS. > > > > I replaced the drive and had the same thing straight away. Replaced > > with another drive and took the unit down to basics including a RAM > > check and rebuilt it. It's been running until today... same thing. > > > > On the messages screen I see at just gone 00:00 this morning... > > Sense Key: aborted command > > Vendor Gen-ATA error code 0x3 > > Rpool has encountered an uncorrectable IO failure and has been > > suspended. Zpool clear will be required before the pool can be > > written > > to > > > > The rpool message repeated at 00:05, 00:22, 02:54 and at 08:54 > > ...and > > it's now stopped responding. > > > > So... I've now got no idea where to take this. On the face of it, > > the > > hardware is fine... on the face of it. However, there's something > > wrong > > here and given the history, I'm not sure where to start. > > > > Part of me thinks I'm looking at hardware... > > > > Power? Hmmm... why does the SSD fail instead of the spinning rust > > drives? > > > > Yet another hard drive? This is number three now, throwing > > suspicion at > > other aspects. > > > > I am reticent to just replace with another hard drive and carry on. > > I > > think it may not be to blame here. > > > > Grateful for people's thoughts. > > > > Michelle. > > > > Apparently this can be multiple things… Does your system allow to set > AHCI (SATA) mode from BIOS setup? This may help. > > also see > https://docs.oracle.com/cd/E19253-01/820-7273/ggmsj/index.html < > https://docs.oracle.com/cd/E19253-01/820-7273/ggmsj/index.html> > > In general, you really want to keep away from IDE… > > rgds, > toomas > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Rpool uncorrectable error
I've done workaround number 5 in that document. The system has been working for years, and I'm not sure about the AHCI... will have to look at that as that will require an OS rebuild, I believe. The only thing that's changed before this started happening, was the OS upgrade. But obviously there's hardware degradation over time to deal with. I'll put a backup and then rebuild on the radar for the next few days, and see what other suggestions come up. Michelle. On Sun, 2021-08-29 at 14:56 +0300, Toomas Soome via openindiana-discuss wrote: > > On 29. Aug 2021, at 14:31, Michelle wrote: > > > > I'm sat here, not knowing quite what I'm dealing with. > > > > OI recent build on HP N54L, as you may remember I hit a problem on > > the > > 8th August with rpool encountering an error on the SSD with the OS. > > > > I replaced the drive and had the same thing straight away. Replaced > > with another drive and took the unit down to basics including a RAM > > check and rebuilt it. It's been running until today... same thing. > > > > On the messages screen I see at just gone 00:00 this morning... > > Sense Key: aborted command > > Vendor Gen-ATA error code 0x3 > > Rpool has encountered an uncorrectable IO failure and has been > > suspended. Zpool clear will be required before the pool can be > > written > > to > > > > The rpool message repeated at 00:05, 00:22, 02:54 and at 08:54 > > ...and > > it's now stopped responding. > > > > So... I've now got no idea where to take this. On the face of it, > > the > > hardware is fine... on the face of it. However, there's something > > wrong > > here and given the history, I'm not sure where to start. > > > > Part of me thinks I'm looking at hardware... > > > > Power? Hmmm... why does the SSD fail instead of the spinning rust > > drives? > > > > Yet another hard drive? This is number three now, throwing > > suspicion at > > other aspects. > > > > I am reticent to just replace with another hard drive and carry on. > > I > > think it may not be to blame here. > > > > Grateful for people's thoughts. > > > > Michelle. > > > > Apparently this can be multiple things… Does your system allow to set > AHCI (SATA) mode from BIOS setup? This may help. > > also see > https://docs.oracle.com/cd/E19253-01/820-7273/ggmsj/index.html < > https://docs.oracle.com/cd/E19253-01/820-7273/ggmsj/index.html> > > In general, you really want to keep away from IDE… > > rgds, > toomas > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Rpool uncorrectable error
root@jaguar:/var# iostat -En c6d1 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Model: KINGSTON SA400S Revision: Serial No: 50026B73801BCEA Size: 120.03GB <120033607680 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 c4t0d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: ATA Product: WDC WD60EFRX-68L Revision: 0A82 Serial No: WD-WX21DA84EH0F Size: 6001.18GB <6001175126016 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 Predictive Failure Analysis: 0 c4t2d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: ATA Product: WDC WD60EFRX-68L Revision: 0A82 Serial No: WD-WX51DB880RJ4 Size: 6001.18GB <6001175126016 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 Predictive Failure Analysis: 0 c4t3d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: ATA Product: WDC WD60EFAX-68J Revision: 0A83 Serial No: WD-WX22DC0AUC60 Size: 6001.18GB <6001175126016 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 Predictive Failure Analysis: 0 Output from fmadm faulty only gives the error which happened when I brought up the unit without the tank pool inserted... just in case. That's as far as I know to go. Michelle. On Sun, 2021-08-29 at 12:56 +0100, Michelle wrote: > I have the machine up and running and issued a zpool clear against > rpool. At least as long as it lasts. > > Reading things from ten years ago, I've disabled the microcode... it > seems to be all I can do. I'm not expecting that to be the issue > but... > hey. > > Nothing in /var/log/syslog which gives any clue. Last message was > 23:05 > from sendmail, and then at 11:46 on a failure to gethostbyaddr after > a > reboot. > > > > On Sun, 2021-08-29 at 12:31 +0100, Michelle wrote: > > I'm sat here, not knowing quite what I'm dealing with. > > > > OI recent build on HP N54L, as you may remember I hit a problem on > > the > > 8th August with rpool encountering an error on the SSD with the OS. > > > > I replaced the drive and had the same thing straight away. Replaced > > with another drive and took the unit down to basics including a RAM > > check and rebuilt it. It's been running until today... same thing. > > > > On the messages screen I see at just gone 00:00 this morning... > > Sense Key: aborted command > > Vendor Gen-ATA error code 0x3 > > Rpool has encountered an uncorrectable IO failure and has been > > suspended. Zpool clear will be required before the pool can be > > written > > to > > > > The rpool message repeated at 00:05, 00:22, 02:54 and at 08:54 > > ...and > > it's now stopped responding. > > > > So... I've now got no idea where to take this. On the face of it, > > the > > hardware is fine... on the face of it. However, there's something > > wrong > > here and given the history, I'm not sure where to start. > > > > Part of me thinks I'm looking at hardware... > > > > Power? Hmmm... why does the SSD fail instead of the spinning rust > > drives? > > > > Yet another hard drive? This is number three now, throwing > > suspicion > > at > > other aspects. > > > > I am reticent to just replace with another hard drive and carry on. > > I > > think it may not be to blame here. > > > > Grateful for people's thoughts. > > > > Michelle. > > > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Rpool uncorrectable error
I have the machine up and running and issued a zpool clear against rpool. At least as long as it lasts. Reading things from ten years ago, I've disabled the microcode... it seems to be all I can do. I'm not expecting that to be the issue but... hey. Nothing in /var/log/syslog which gives any clue. Last message was 23:05 from sendmail, and then at 11:46 on a failure to gethostbyaddr after a reboot. On Sun, 2021-08-29 at 12:31 +0100, Michelle wrote: > I'm sat here, not knowing quite what I'm dealing with. > > OI recent build on HP N54L, as you may remember I hit a problem on > the > 8th August with rpool encountering an error on the SSD with the OS. > > I replaced the drive and had the same thing straight away. Replaced > with another drive and took the unit down to basics including a RAM > check and rebuilt it. It's been running until today... same thing. > > On the messages screen I see at just gone 00:00 this morning... > Sense Key: aborted command > Vendor Gen-ATA error code 0x3 > Rpool has encountered an uncorrectable IO failure and has been > suspended. Zpool clear will be required before the pool can be > written > to > > The rpool message repeated at 00:05, 00:22, 02:54 and at 08:54 ...and > it's now stopped responding. > > So... I've now got no idea where to take this. On the face of it, the > hardware is fine... on the face of it. However, there's something > wrong > here and given the history, I'm not sure where to start. > > Part of me thinks I'm looking at hardware... > > Power? Hmmm... why does the SSD fail instead of the spinning rust > drives? > > Yet another hard drive? This is number three now, throwing suspicion > at > other aspects. > > I am reticent to just replace with another hard drive and carry on. I > think it may not be to blame here. > > Grateful for people's thoughts. > > Michelle. > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Rpool uncorrectable error
I'm sat here, not knowing quite what I'm dealing with. OI recent build on HP N54L, as you may remember I hit a problem on the 8th August with rpool encountering an error on the SSD with the OS. I replaced the drive and had the same thing straight away. Replaced with another drive and took the unit down to basics including a RAM check and rebuilt it. It's been running until today... same thing. On the messages screen I see at just gone 00:00 this morning... Sense Key: aborted command Vendor Gen-ATA error code 0x3 Rpool has encountered an uncorrectable IO failure and has been suspended. Zpool clear will be required before the pool can be written to The rpool message repeated at 00:05, 00:22, 02:54 and at 08:54 ...and it's now stopped responding. So... I've now got no idea where to take this. On the face of it, the hardware is fine... on the face of it. However, there's something wrong here and given the history, I'm not sure where to start. Part of me thinks I'm looking at hardware... Power? Hmmm... why does the SSD fail instead of the spinning rust drives? Yet another hard drive? This is number three now, throwing suspicion at other aspects. I am reticent to just replace with another hard drive and carry on. I think it may not be to blame here. Grateful for people's thoughts. Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Memory Test in OpenIndiana
Apparently there is an issue with later versions of Memtest86 on the N54L motherboards. Even a direct download of Memtest on a live USB key failed to boot. However, I found a thread that suggested getting the Ubuntu 18.04 version and running Memtest on that version. So far... so running. On Sun, 2021-08-08 at 15:08 +0100, Michelle wrote: > I've got two N54L but trying to run MemTest through Ubuntu > installation > media results in hangs on most memory chips which might be something > to > do with the custom BIOS. I'm not sure. > > I've seen nothing in messages to indicate memory issues, but I really > don't know what to do now. If I can't test the RAM, I can't rule it > out. > > Given the BIOS issue then testing under OpenIndiana would be the best > option, but I can't find out how to do it. > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Memory Test in OpenIndiana
I've got two N54L but trying to run MemTest through Ubuntu installation media results in hangs on most memory chips which might be something to do with the custom BIOS. I'm not sure. I've seen nothing in messages to indicate memory issues, but I really don't know what to do now. If I can't test the RAM, I can't rule it out. Given the BIOS issue then testing under OpenIndiana would be the best option, but I can't find out how to do it. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] N54L any known issues?
Bit of an odd one this, the last couple of days have been a bit of drama. To cut to the long story, the OS drive (an SSD) failed... in the middle of dealing with the failed "tank" pool I replaced it and built the OS again on a new drive, but that's failed as well in less than a day with checksum errors. I've got a memory check on the list of things to do, but beyond that I'm at a bit of a blank. I'm also planning to replace the new SSD with another SSD... but if that fails in short order as well, then I'm going to be pointing the finger at the server itself. The copying from backup drive is continuing, although I can't log into the server to check anything as any logon prompt just hangs... which is understandable. I know that the other N54L was failing drives on one of the SATA channels and my conclusion is that it wasn't the drives at fault... so I'm wondering if anyone has any knowledge of SATA channel breakdowns on the N54L units. It's regularly cleaned of excess dust, etc. so it's not the kind of a box that's left to gather dust and rot until failure. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] rpool mirror disk problems when moved to different slot
Ah, if you're moving a live rpool, that's a bit more complex. I'd have destroyed the mirror, moved the drive and then re-mirrored... but only if I'd put another boot on the other drive. After all, if you've mirrored your rpool, but there's no boot on the mirror drive, then it will still not boot the mirror drive. On Fri, 2021-08-06 at 12:54 +0200, Stephan Althaus wrote: > On 8/6/21 12:47 PM, Michelle wrote: > > Did you export the zpool before moving the drives and then import > > again > > after? > > > > If I don't export before I move things, then I wouldn't expect ZFS > > to > > be able to track it automatically. > > > > > > > > On Fri, 2021-08-06 at 11:19 +0200, Stephan Althaus wrote: > > > Hello! > > > > > > i just added a new disk to my system, which i put into the slot > > > where > > > a > > > mirror disk of my rpool resided. > > > > > > The mirror disk of my rpool went to a other slot, so the device > > > name > > > was > > > changed c27t1d0 -> c27t0d0 > > > > > > Booting into my OI system, the zpool was found as the first > > > mirror > > > disk > > > was there, > > > but the second partition was not found. > > > > > > > > > Is there a chance that the rpool mount procedure could try harder > > > to > > > find the mirrors, even when moved ?? > > > > > > > > > Yes i know this may be a rare case :-/ > > > > > > > > > Greetings, > > > > > > Stephan > > > > > > > > > ___ > > > openindiana-discuss mailing list > > > openindiana-discuss@openindiana.org > > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > Hi! > > This is my root pool, is it exported on shutdown? i would think so.. > > I think, if i move the mirror drive to a different slot and leave the > origin-slot empty, it could work.. > > I should have imported and exported the pool with a illumos-live- > USB, > _before_ trying to boot the installed system with the change. > > i had no live-medium at hand so i just gave it a try.. > > Stephan > > BTW, the mirrored pool is up and working: > > $ zpool status rpool > pool: rpool > state: ONLINE > scan: resilvered 194G in 0 days 00:50:39 with 0 errors on Fri Aug > 6 > 12:04:20 2021 > config: > > NAME STATE READ WRITE CKSUM > rpoolONLINE 0 0 0 > mirror-0 ONLINE 0 0 0 > c9t002538B471B9EFD3d0s8 ONLINE 0 0 0 > c27t0d0s0ONLINE 0 0 0 > > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] rpool mirror disk problems when moved to different slot
Did you export the zpool before moving the drives and then import again after? If I don't export before I move things, then I wouldn't expect ZFS to be able to track it automatically. On Fri, 2021-08-06 at 11:19 +0200, Stephan Althaus wrote: > Hello! > > i just added a new disk to my system, which i put into the slot where > a > mirror disk of my rpool resided. > > The mirror disk of my rpool went to a other slot, so the device name > was > changed c27t1d0 -> c27t0d0 > > Booting into my OI system, the zpool was found as the first mirror > disk > was there, > but the second partition was not found. > > > Is there a chance that the rpool mount procedure could try harder to > find the mirrors, even when moved ?? > > > Yes i know this may be a rare case :-/ > > > Greetings, > > Stephan > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Errors without errors
OK ... to update... Apparently I did the update on 30th August. Previous installation was some time in May. I used beadm to revert to the previous boot environment, then rebooted and destroyed the August environment. The system then faulted rpool saying it needed zpool clear and refused to boot, just hanging at the "Hipster" load screen. I used option 3 for the loader prompt and couldn't find anything in there that would help, but for the giggles I told it to boot anyway... and it did. So... exactly what happened there, I don't know. It resilvered the data set, which didn't take long at all... minutes... so I've got a scrub running and so far, no errors on /var/adm/messages Fingers crossed. Scheduled to take 16 hours. I'm keeping an eye on the scrub status and the messages file and we'll see what happens. Michelle. On Thu, 2021-08-05 at 10:46 -0300, Till Wegmueller wrote: > Hi Michelle > > For User home files you will need a backup anyway. For system > Consistency you can use `pkg fix` to restore the system image to a > known > state in a new Boot environment. > > Greetings > Till > > On 05.08.21 05:14, Toomas Soome via openindiana-discuss wrote: > > > > > On 5. Aug 2021, at 11:11, Michelle wrote: > > > > > > I removed the drive in order to a backup before I start messing > > > around > > > with things, which is why it isn't in the iostat. The backup will > > > take > > > probably until early evening. > > > > > > This is what happened from messages around that time. Almost > > > looks like > > > whatever happened, it rebooted. > > > > > > > From those, I’d say, you need to replace that disk. > > > > rgds, > > toomas > > > > > Aug 5 01:55:01 jaguar smbd[601]: [ID 617204 daemon.error] Can't > > > get > > > SID for ID=0 type=1, status=-9977 > > > Aug 5 01:58:00 jaguar ahci: [ID 296163 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 has task file error > > > Aug 5 01:58:00 jaguar ahci: [ID 687168 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 is trying to do error recovery > > > Aug 5 01:58:00 jaguar ahci: [ID 693748 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 task_file_status = 0x4041 > > > Aug 5 01:58:00 jaguar ahci: [ID 657156 kern.warning] WARNING: > > > ahci0: > > > error recovery for port 3 succeed > > > Aug 5 01:58:09 jaguar ahci: [ID 296163 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 has task file error > > > Aug 5 01:58:09 jaguar ahci: [ID 687168 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 is trying to do error recovery > > > Aug 5 01:58:09 jaguar ahci: [ID 693748 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 task_file_status = 0x4041 > > > Aug 5 01:58:09 jaguar ahci: [ID 657156 kern.warning] WARNING: > > > ahci0: > > > error recovery for port 3 succeed > > > Aug 5 02:00:15 jaguar ahci: [ID 296163 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 has task file error > > > Aug 5 02:00:15 jaguar ahci: [ID 687168 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 is trying to do error recovery > > > Aug 5 02:00:15 jaguar ahci: [ID 693748 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 task_file_status = 0x4041 > > > Aug 5 02:00:16 jaguar ahci: [ID 657156 kern.warning] WARNING: > > > ahci0: > > > error recovery for port 3 succeed > > > Aug 5 02:00:20 jaguar ahci: [ID 296163 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 has task file error > > > Aug 5 02:00:20 jaguar ahci: [ID 687168 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 is trying to do error recovery > > > Aug 5 02:00:20 jaguar ahci: [ID 693748 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 task_file_status = 0x4041 > > > Aug 5 02:00:20 jaguar ahci: [ID 657156 kern.warning] WARNING: > > > ahci0: > > > error recovery for port 3 succeed > > > Aug 5 02:00:24 jaguar ahci: [ID 296163 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 has task file error > > > Aug 5 02:00:24 jaguar ahci: [ID 687168 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 is trying to do error recovery > > > Aug 5 02:00:24 jaguar ahci: [ID 693748 kern.warning] WARNING: > > > ahci0: > > > ahci port 3 task_file_status = 0x4041 > > > Aug 5 02:00:24 jaguar ahci: [ID
Re: [OpenIndiana-discuss] Errors without errors
I removed the drive in order to a backup before I start messing around with things, which is why it isn't in the iostat. The backup will take probably until early evening. This is what happened from messages around that time. Almost looks like whatever happened, it rebooted. Aug 5 01:55:01 jaguar smbd[601]: [ID 617204 daemon.error] Can't get SID for ID=0 type=1, status=-9977 Aug 5 01:58:00 jaguar ahci: [ID 296163 kern.warning] WARNING: ahci0: ahci port 3 has task file error Aug 5 01:58:00 jaguar ahci: [ID 687168 kern.warning] WARNING: ahci0: ahci port 3 is trying to do error recovery Aug 5 01:58:00 jaguar ahci: [ID 693748 kern.warning] WARNING: ahci0: ahci port 3 task_file_status = 0x4041 Aug 5 01:58:00 jaguar ahci: [ID 657156 kern.warning] WARNING: ahci0: error recovery for port 3 succeed Aug 5 01:58:09 jaguar ahci: [ID 296163 kern.warning] WARNING: ahci0: ahci port 3 has task file error Aug 5 01:58:09 jaguar ahci: [ID 687168 kern.warning] WARNING: ahci0: ahci port 3 is trying to do error recovery Aug 5 01:58:09 jaguar ahci: [ID 693748 kern.warning] WARNING: ahci0: ahci port 3 task_file_status = 0x4041 Aug 5 01:58:09 jaguar ahci: [ID 657156 kern.warning] WARNING: ahci0: error recovery for port 3 succeed Aug 5 02:00:15 jaguar ahci: [ID 296163 kern.warning] WARNING: ahci0: ahci port 3 has task file error Aug 5 02:00:15 jaguar ahci: [ID 687168 kern.warning] WARNING: ahci0: ahci port 3 is trying to do error recovery Aug 5 02:00:15 jaguar ahci: [ID 693748 kern.warning] WARNING: ahci0: ahci port 3 task_file_status = 0x4041 Aug 5 02:00:16 jaguar ahci: [ID 657156 kern.warning] WARNING: ahci0: error recovery for port 3 succeed Aug 5 02:00:20 jaguar ahci: [ID 296163 kern.warning] WARNING: ahci0: ahci port 3 has task file error Aug 5 02:00:20 jaguar ahci: [ID 687168 kern.warning] WARNING: ahci0: ahci port 3 is trying to do error recovery Aug 5 02:00:20 jaguar ahci: [ID 693748 kern.warning] WARNING: ahci0: ahci port 3 task_file_status = 0x4041 Aug 5 02:00:20 jaguar ahci: [ID 657156 kern.warning] WARNING: ahci0: error recovery for port 3 succeed Aug 5 02:00:24 jaguar ahci: [ID 296163 kern.warning] WARNING: ahci0: ahci port 3 has task file error Aug 5 02:00:24 jaguar ahci: [ID 687168 kern.warning] WARNING: ahci0: ahci port 3 is trying to do error recovery Aug 5 02:00:24 jaguar ahci: [ID 693748 kern.warning] WARNING: ahci0: ahci port 3 task_file_status = 0x4041 Aug 5 02:00:24 jaguar ahci: [ID 657156 kern.warning] WARNING: ahci0: error recovery for port 3 succeed Aug 5 02:00:24 jaguar ahci: [ID 811322 kern.info] NOTICE: ahci0: ahci_tran_reset_dport port 3 reset device Aug 5 02:00:29 jaguar ahci: [ID 296163 kern.warning] WARNING: ahci0: ahci port 3 has task file error Aug 5 02:00:29 jaguar ahci: [ID 687168 kern.warning] WARNING: ahci0: ahci port 3 is trying to do error recovery Aug 5 02:00:29 jaguar ahci: [ID 693748 kern.warning] WARNING: ahci0: ahci port 3 task_file_status = 0x4041 Aug 5 02:00:29 jaguar ahci: [ID 657156 kern.warning] WARNING: ahci0: error recovery for port 3 succeed Aug 5 02:00:34 jaguar ahci: [ID 296163 kern.warning] WARNING: ahci0: ahci port 3 has task file error Aug 5 02:00:34 jaguar ahci: [ID 687168 kern.warning] WARNING: ahci0: ahci port 3 is trying to do error recovery Aug 5 02:00:34 jaguar ahci: [ID 693748 kern.warning] WARNING: ahci0: ahci port 3 task_file_status = 0x4041 Aug 5 02:00:34 jaguar ahci: [ID 657156 kern.warning] WARNING: ahci0: error recovery for port 3 succeed Aug 5 02:00:38 jaguar ahci: [ID 296163 kern.warning] WARNING: ahci0: ahci port 3 has task file error Aug 5 02:00:38 jaguar ahci: [ID 687168 kern.warning] WARNING: ahci0: ahci port 3 is trying to do error recovery Aug 5 02:00:38 jaguar ahci: [ID 693748 kern.warning] WARNING: ahci0: ahci port 3 task_file_status = 0x4041 Aug 5 02:00:38 jaguar ahci: [ID 657156 kern.warning] WARNING: ahci0: error recovery for port 3 succeed Aug 5 02:00:53 jaguar fmd: [ID 377184 daemon.error] SUNW-MSG-ID: ZFS- 8000-FD, TYPE: Fault, VER: 1, SEVERITY: Major Aug 5 02:00:53 jaguar EVENT-TIME: Thu Aug 5 02:00:53 UTC 2021 Aug 5 02:00:53 jaguar PLATFORM: ProLiant-MicroServer, CSN: 5C7351P4L9, HOSTNAME: jaguar Aug 5 02:00:53 jaguar SOURCE: zfs-diagnosis, REV: 1.0 On Thu, 2021-08-05 at 11:03 +0300, Toomas Soome via openindiana-discuss wrote: > > On 5. Aug 2021, at 10:52, Michelle wrote: > > > > Thanks for this. So I'm possibly better off rolling back the OS > > snapshot after my backup has finished? > > maybe, maybe not. first of all, I have no idea to what point the > rollback would be. > > secondly; the system has seen some errors, at this time, the fault > is, it does not tell us if those were checksum errors or something > else, and it seems to me, it is something else. > > and this is why: if you look on your zpool output, you see report > about c6t3d0, but iostat -En below, it does not include c6t3d0. It > seems to be missing. >
Re: [OpenIndiana-discuss] Errors without errors
Thanks for this. So I'm possibly better off rolling back the OS snapshot after my backup has finished? I have removed the drive for the moment, and am running a backup. Just in case :-) mich@jaguar:~$ iostat -En c5d1 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Model: INTEL SSDSA2M04 Revision: Serial No: CVGB949301PC040 Size: 40.02GB <40019116032 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 c6t1d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: ATA Product: WDC WD40EZRZ-00G Revision: 0A80 Serial No: WD-WCC7K5UK24LJ Size: 4000.79GB <4000787030016 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 Predictive Failure Analysis: 0 c6t0d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: ATA Product: WDC WD60EFRX-68L Revision: 0A82 Serial No: WD-WX21DA84EH0F Size: 6001.18GB <6001175126016 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 Predictive Failure Analysis: 0 c6t2d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: ATA Product: WDC WD60EFRX-68L Revision: 0A82 Serial No: WD-WX51DB880RJ4 Size: 6001.18GB <6001175126016 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 Predictive Failure Analysis: 0 --- -- -- --- TIMEEVENT-ID MSG- ID SEVERITY --- -- -- --- Aug 05 02:00:53 c5934fd6-5f4b-409e-b0f8-8f44ea8f99c4 ZFS-8000- FDMajor Host: jaguar Platform: ProLiant-MicroServer Chassis_id : 5C7351P4L9 Product_sn : Fault class : fault.fs.zfs.vdev.io Affects : zfs://pool=jaguar/vdev=740c01ae0d3c3109 faulted and taken out of service Problem in : zfs://pool=jaguar/vdev=740c01ae0d3c3109 faulted and taken out of service Description : The number of I/O errors associated with a ZFS device exceeded acceptable levels. Refer to http://illumos.org/msg/ZFS-8000-FD for more information. Response: The device has been offlined and marked as faulted. An attempt will be made to activate a hot spare if available. Impact : Fault tolerance of the pool may be compromised. Action : Run 'zpool status -x' and replace the bad device. On Thu, 2021-08-05 at 10:22 +0300, Toomas Soome via openindiana-discuss wrote: > > On 5. Aug 2021, at 09:35, Michelle wrote: > > > > Hi Folks, > > > > About a month ago I updated my Hipster... > > SunOS jaguar 5.11 illumos-ca706442e6 i86pc i386 i86pc > > > > This morning it was absolutely crawling. Couldn't even connect via > > SSH > > and had to bounce the box. > > > > It was reporting a drive as faulted, but didn't give any numbers... > > everything was 0. I'm now not sure what happened and whether the > > drive > > is good, or whether I should roll back the OS. > > > > (and the drive WD Red 6TB (not shingle) went out of warrantee a > > week > > ago. How about that, eh?) > > > > Grateful for any opinions please. > > > > Thu 5 Aug 04:00:01 UTC 2021 > > NAME SIZE ALLOC FREE CKPOINT EXPANDSZ FRAGCAP DEDUP > > HE > > ALTH ALTROOT > > lion 5.45T 5.28T 176G- - 4%96% 1.00x > > DEGR > > ADED - > > pool: jaguar > > state: DEGRADED > > status: One or more devices are faulted in response to persistent > > errors. > > Sufficient replicas exist for the pool to continue functioning > > in a > > degraded state. > > action: Replace the faulted device, or use 'zpool clear' to mark > > the > > device > > repaired. > > scan: scrub in progress since Thu Aug 5 00:00:00 2021 > > 6.00T scanned at 428M/s, 5.02T issued at 358M/s, 7.90T total > > 1M repaired, 63.59% done, 0 days 02:20:17 to go > > config: > > NAMESTATE READ WRITE CKSUM > > jaguar DEGRADED 0 0 0 > > raidz1-0 DEGRADED 0 0 0 > > c6t0d0 ONLINE 0 0 0 > > c6t2d0 ONLINE 0 0 0 > > c6t3d0 FAULTED 0 0 0 too many > > errors (repairing) > > > > Can you postoutput from: > iostat -En > fmadm faulty > > in any case, there definitely is bug about error reporting - counters > are zero while “too many errors” is reported. > > rgds, > toomas > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Errors without errors
Hi Folks, About a month ago I updated my Hipster... SunOS jaguar 5.11 illumos-ca706442e6 i86pc i386 i86pc This morning it was absolutely crawling. Couldn't even connect via SSH and had to bounce the box. It was reporting a drive as faulted, but didn't give any numbers... everything was 0. I'm now not sure what happened and whether the drive is good, or whether I should roll back the OS. (and the drive WD Red 6TB (not shingle) went out of warrantee a week ago. How about that, eh?) Grateful for any opinions please. Thu 5 Aug 04:00:01 UTC 2021 NAME SIZE ALLOC FREE CKPOINT EXPANDSZ FRAGCAP DEDUPHE ALTH ALTROOT lion 5.45T 5.28T 176G- - 4%96% 1.00x DEGR ADED - pool: jaguar state: DEGRADED status: One or more devices are faulted in response to persistent errors. Sufficient replicas exist for the pool to continue functioning in a degraded state. action: Replace the faulted device, or use 'zpool clear' to mark the device repaired. scan: scrub in progress since Thu Aug 5 00:00:00 2021 6.00T scanned at 428M/s, 5.02T issued at 358M/s, 7.90T total 1M repaired, 63.59% done, 0 days 02:20:17 to go config: NAMESTATE READ WRITE CKSUM jaguar DEGRADED 0 0 0 raidz1-0 DEGRADED 0 0 0 c6t0d0 ONLINE 0 0 0 c6t2d0 ONLINE 0 0 0 c6t3d0 FAULTED 0 0 0 too many errors (repairing) ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] read/write hard drive parameters
Hi Tony, Many thanks for the links. I can fortunately report that "format" reports correctly for the 8tb drives. The SMR drives stopped at 6TB, and WD faced a bit of criticism over their decision to ship those in their RED lineup, so I'm not expecting these to be the issue in this case as these are 8Tb. Michelle. On Wed, 2021-05-12 at 06:14 +, Tony Brian Albers wrote: > On Wed, 2021-05-12 at 06:43 +0100, Michelle wrote: > > I'm having awful luck with WD 8Tb red drives. > > > > Mirror configuration, ZFS is failing the drives on sustained > > writes. > > RAIDZ seems to be more forgiving. > > > > I've sent one drive back and have two more under RMA, but so far my > > crude testing seems to indicate that this is a configuration issue > > generic to the model, rather than an out and out faulty drive. > > > > Someone else said (not on here) has said they had to tweek the > > write > > timeouts on the drives to stop them erroring. > > > > So the trick is... how do I address the drive parameters under > > OpenIndiana please? Does anyone have a link to a how-to please? > > > > Michelle. > > > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > Seems like this danish carpenter(!) knows what the issue is: > https://ronn-bundgaard.dk/blog/freenas-zfs-does-not-like-the-new-wd-red-smr-drives/ > > This might be of some value: > https://unix.stackexchange.com/questions/447318/how-to-force-omnios-illumos-format-properly-recognize-disk-geometry > > And here are mentioned some mitigation techniques: > https://www.truenas.com/community/threads/update-wd-red-smr-drive-compatibility-with-zfs.88413/ > > HTH > > /tony (who is also danish, and a combat engineer/sapper) > ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] read/write hard drive parameters
I'm having awful luck with WD 8Tb red drives. Mirror configuration, ZFS is failing the drives on sustained writes. RAIDZ seems to be more forgiving. I've sent one drive back and have two more under RMA, but so far my crude testing seems to indicate that this is a configuration issue generic to the model, rather than an out and out faulty drive. Someone else said (not on here) has said they had to tweek the write timeouts on the drives to stop them erroring. So the trick is... how do I address the drive parameters under OpenIndiana please? Does anyone have a link to a how-to please? Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Partitions for co-exist NAS drive
I just followed the text install on the 2021.04 GUi live image, and it created a single Rpool for the whole disk. I'm not sure where I'm going to go from here. Need to have a sit and think. Thank you very much for all your efforts on this. I really do appreciate it. Michelle. On Sat, 2021-05-01 at 16:23 +, Reginald Beardsley via openindiana- discuss wrote: > I just installed the 2021.04.30 text ISO on my Z400 test system. The > disk had a pool in a single slice which I imported and destroyed. Had > I not done that it would have named the pool rpool1. I left the disk > label alone. This was from testing manually creating a single slice > and installing into that using the F5 option which doesn't create > dump and swap on the expectation they already exist. > > The installer still has the usual spurious messages about only using > 2 TB. I selected an EFI label and did the install. > > That created a 256 MB s0 slice beginning in sector 256 with the rest > of the disk in s1. > > "zfs list" shows the pool has 4.39 TB available which is correct for > having used the entire disk. > > If you're going to run a 2 disk mirror with the 6 TB drives I'd > install to a single pool using the text installer. Assuming it works > properly. When the 8 TB drive returns, copy the data and then create > a matching label and add the drive to the pool to create a mirror. > > Unfortunately we are a tad shy on current documentation. > > Reg > > On Saturday, May 1, 2021, 10:15:53 AM CDT, Michelle < > miche...@msknight.com> wrote: > > Actually, I was wrong. It was the 2021.04 text install USB that I > was > using. I was using the text install because it's just a file server. > I > don't need X. > > Well, the situation I find myself in is as follows... > > A pair of 6tb reds, one of which has data on that I can't get off > until > one of my 8tb has come back from RMA > > So I thought I'd use the other 6tb to get this sorted, and by the > time > the other 6tb is ready to be put in, I would be able to install it > and > add it in. > > Going via the GUI installer, I'm wondering whether I'll be able to do > that. > > This is why I was following your script, but when it came to step 4, > the solaris partition was only 2tb max. However, I've just noted that > I > wasn't using the -e at the end of the format command, so after I've > made dinner I'll read up on that switch in the hope that this is what > will enable me to create the larger Solaris partition. > > Michelle. > > > On Sat, 2021-05-01 at 14:04 +, Reginald Beardsley via > openindiana- > discuss wrote: > > Michelle, > > > > What disks are you trying to use? If they are different sizes the > > recipe gets a bit more complex, but it's quite possible to get any > > arrangement for which you have a sufficient number of disks. Having > > spent 50 hours over 5 days battling the 2021.04.05 install I've had > > *lots* of practice installing both 2021.04.05 and 2020.10. > > > > Incidentally the failure of 2021.04.05 to run the nVIDIA driver is > > because it installed the wrong driver for that system. > > > > Reg > > > > > > > > > > On Saturday, May 1, 2021, 08:16:07 AM CDT, Michelle < > > miche...@msknight.com> wrote: > > > > > > OK - I give in. > > > > I've tried various combinations and I just can't get this into a > > configuration where I can get things installed. > > > > I need a step by step guide please, because I'm lost. I don't know > > what > > options to choose in the installer to not avoid it wanting to wipe > > whatever combinations of partitions I've set. > > > > Michelle. > > > > > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Partitions for co-exist NAS drive
Actually, I was wrong. It was the 2021.04 text install USB that I was using. I was using the text install because it's just a file server. I don't need X. Well, the situation I find myself in is as follows... A pair of 6tb reds, one of which has data on that I can't get off until one of my 8tb has come back from RMA So I thought I'd use the other 6tb to get this sorted, and by the time the other 6tb is ready to be put in, I would be able to install it and add it in. Going via the GUI installer, I'm wondering whether I'll be able to do that. This is why I was following your script, but when it came to step 4, the solaris partition was only 2tb max. However, I've just noted that I wasn't using the -e at the end of the format command, so after I've made dinner I'll read up on that switch in the hope that this is what will enable me to create the larger Solaris partition. Michelle. On Sat, 2021-05-01 at 14:04 +, Reginald Beardsley via openindiana- discuss wrote: > Michelle, > > What disks are you trying to use? If they are different sizes the > recipe gets a bit more complex, but it's quite possible to get any > arrangement for which you have a sufficient number of disks. Having > spent 50 hours over 5 days battling the 2021.04.05 install I've had > *lots* of practice installing both 2021.04.05 and 2020.10. > > Incidentally the failure of 2021.04.05 to run the nVIDIA driver is > because it installed the wrong driver for that system. > > Reg > > > > > On Saturday, May 1, 2021, 08:16:07 AM CDT, Michelle < > miche...@msknight.com> wrote: > > > OK - I give in. > > I've tried various combinations and I just can't get this into a > configuration where I can get things installed. > > I need a step by step guide please, because I'm lost. I don't know > what > options to choose in the installer to not avoid it wanting to wipe > whatever combinations of partitions I've set. > > Michelle. > > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Partitions for co-exist NAS drive
Thanks. I'll give that a go. I'm using the text installer 2020.10 and on that, I can't create a Solaris partition greater than 2tb. I've got the GUI live image here somewhere. Michelle. On Sat, 2021-05-01 at 13:50 +, Reginald Beardsley via openindiana- discuss wrote: > What release are you trying to install? I last did this using > 2021.04.05, but I can easily verify the behavior with 2020.10. > > I *strongly* recommend that you put all the disks you plan to use in > the machine and run the text-installer from the Live Desktop. Select > F2 on the first screen. When you get to the disk list, select all the > disks. Make sure you don't have a USB stick plugged in. I wiped out > mine by accident. Selecting multiple disks triggers the appearance of > options to create a mirror, RAIDZ1 or RAIDZ2 depending on whether you > have 2, 3 or 4 disks. Just make sure you select an EFI label, not MBR > > Let it wipe out the partitions and create the 250 MB slice and the > rest of disk slice. That *is* the best practice today. The mirrored > root pool slice is a bodge for releases which can't boot from RAIDZ. > > If for some reason you want to do something different this is the > step by step recipe. It will give you exactly whatever you set up > with format(1m). Because the F5 option is expecting to be used on an > existing bootable pool it doesn't create dump and swap, so you have > to do that by hand. The absence of /reconfigure was what beat me up > for a week before I figured it out quite by accident. > > All this is done in a terminal window after doing "sudo /bin/su" as > jack. > > > format -e > # select disk > > fdisk > # create Solaris partition for entire drive and commit to disk > > partition > # create the desired slices > > label > # write an EFI label > > quit > > verify > > quit > > You should now have a Sun EFI label with 9 slices. 8 is set by > format(1m) and can't be changed. The other two should be the ones you > created. You will need to do this for each disk and then check using > prtvtoc(1m) that they all match. The disks don't need to match, but > the slice sizes should. > > I should note that format(1m) will not allow deleting a slice by > setting the start and end to 0 easily. I have managed to do it, but > I'm not sure what the correct incantation is. It gives "0 is out of > range" messages most of the time. > > In the first text installer screen chose F5, in the 2nd screen select > the slice you want to use. Continue with the install. When It > completes reboot to single user. > > zfs create -V rpool/dump rpool > zfs create -V rpool/swap rpool > dumpadm -d /dev/zvol/dsk/rpool/dump > swap -a /dev/zvol/dsk/rpool/swap > touch /reconfigure > init 6 > > Reg On Saturday, May 1, 2021, 08:16:07 AM CDT, Michelle < > miche...@msknight.com> wrote: > > OK - I give in. > > I've tried various combinations and I just can't get this into a > configuration where I can get things installed. > > I need a step by step guide please, because I'm lost. I don't know > what > options to choose in the installer to not avoid it wanting to wipe > whatever combinations of partitions I've set. > > Michelle. > > > On Sat, 2021-05-01 at 15:42 +0300, Toomas Soome via openindiana- > discuss > wrote: > > > On 1. May 2021, at 15:30, Michelle wrote: > > > > > > That's where I'm becoming unstuck. > > > > > > A Solaris 2 partition will only see the first 2Tb. > > > > > > There hangs my first problem. > > > > > > If I try and create any other partition it gives me the warning > > > about > > > the 2TB limit and if I then try and create the EFI partition, it > > > won't > > > co-exist with anything and wants to wipe the whole disk again. > > > > > > Michelle. > > > > MBR and GPT can not co-exist. On this disk, you need GPT and this > > means, MBR partitions will be removed. > > > > Rgds, > > Toomas > > > > > > On Sat, 2021-05-01 at 12:21 +, Reginald Beardsley via > > > > openindiana- > > > > discuss wrote: > > > > > > > > I just went through several iterations of this, and like you > > > > the > > > > last > > > > time I had done it was long ago. The following is based on > > > > 2021.04.05. > > > > > > > > Large disks require an EFI or GPT label. The gparted program > > > > creates > > > &g
Re: [OpenIndiana-discuss] Partitions for co-exist NAS drive
OK - I give in. I've tried various combinations and I just can't get this into a configuration where I can get things installed. I need a step by step guide please, because I'm lost. I don't know what options to choose in the installer to not avoid it wanting to wipe whatever combinations of partitions I've set. Michelle. On Sat, 2021-05-01 at 15:42 +0300, Toomas Soome via openindiana-discuss wrote: > > On 1. May 2021, at 15:30, Michelle wrote: > > > > That's where I'm becoming unstuck. > > > > A Solaris 2 partition will only see the first 2Tb. > > > > There hangs my first problem. > > > > If I try and create any other partition it gives me the warning > > about > > the 2TB limit and if I then try and create the EFI partition, it > > won't > > co-exist with anything and wants to wipe the whole disk again. > > > > Michelle. > > MBR and GPT can not co-exist. On this disk, you need GPT and this > means, MBR partitions will be removed. > > Rgds, > Toomas > > > > > > On Sat, 2021-05-01 at 12:21 +, Reginald Beardsley via > > > openindiana- > > > discuss wrote: > > > > > > I just went through several iterations of this, and like you the > > > last > > > time I had done it was long ago. The following is based on > > > 2021.04.05. > > > > > > Large disks require an EFI or GPT label. The gparted program > > > creates > > > 128 slices which is a bit much. format(1m) will also write an EFI > > > label which is usable with large disks. > > > > > > > format -e > > > # select disk > > > > fdisk > > > # create Solaris partition for entire drive and commit to disk > > > > partition > > > # create the desired slices and write an EFI label > > > > quit > > > > verify > > > > quit > > > > > > You should now have a Sun EFI label with 9 slices. 8 is set by > > > format(1m) and can't be changed. The other two should be the ones > > > you > > > created. > > > > > > In the first text installer screen chose F5, in the 2nd screen > > > select > > > the slice you want to use. Continue with the install. When It > > > completes reboot to single user. > > > > > > zfs create -V rpool/dump rpool > > > zfs create -V rpool/swap rpool > > > dumpadm -d /dev/zvol/dsk/rpool/dump > > > swap -a /dev/zvol/dsk/rpool/swap > > > touch /reconfigure > > > init 6 > > > > > > You should now come up with rpool in the 100 GB slice. > > > > > > That said, we can boot from RAIDZ now. The text-install on the > > > Desktop live image will let you create a mirror, RAIDZ1 or RAIDZ2 > > > and > > > will take care of all the label stuff. Despite the statement that > > > it > > > will only use 2 TB, it in fact uses the entire disk. > > > > > > It creates a 250 MB s0 slice and the rest of the disk in s1. The > > > 250 > > > MB slice is labeled "System", but I've not seen any explanation > > > of > > > it. I've also created RAIDZ2 pools by hand and used F5 to install > > > into them. F5 appears to be intended to install into a new BE in > > > an > > > existing pool, hence the need to set up dump and swap by hand. > > > > > > Ultimately I decided I didn't care about 1 GB of unused space in > > > 16 > > > TB of space. So I just went with the text-install created RAIDZ2 > > > pool. The reconfigure on the first boot after the install is > > > critical > > > to getting 2021.04.05 up properly. > > > > > > Reg > > > On Saturday, May 1, 2021, 02:57:23 AM CDT, Michelle < > > > miche...@msknight.com> wrote: > > > > > > OK - I appear to be well out of touch. > > > > > > I booted the installer and went into prompt. > > > > > > Used format (only 1 x 6TB drive in the machine at this point) to > > > create > > > a new Solaris 2 partition table and then fdisk'd an all free hog > > > to > > > partition 1, giving partition 0 100gig. > > > > > > I noticed that it must have gone on for 40 odd partitions, and > > > also > > > there was none of the usual backup and reserved partitions for 2, > > > 8 > > > and > > > 9 as I saw before. > > > > > >
Re: [OpenIndiana-discuss] Partitions for co-exist NAS drive
That's where I'm becoming unstuck. A Solaris 2 partition will only see the first 2Tb. There hangs my first problem. If I try and create any other partition it gives me the warning about the 2TB limit and if I then try and create the EFI partition, it won't co-exist with anything and wants to wipe the whole disk again. Michelle. On Sat, 2021-05-01 at 12:21 +, Reginald Beardsley via openindiana- discuss wrote: > > I just went through several iterations of this, and like you the last > time I had done it was long ago. The following is based on > 2021.04.05. > > Large disks require an EFI or GPT label. The gparted program creates > 128 slices which is a bit much. format(1m) will also write an EFI > label which is usable with large disks. > > > format -e > # select disk > > fdisk > # create Solaris partition for entire drive and commit to disk > > partition > # create the desired slices and write an EFI label > > quit > > verify > > quit > > You should now have a Sun EFI label with 9 slices. 8 is set by > format(1m) and can't be changed. The other two should be the ones you > created. > > In the first text installer screen chose F5, in the 2nd screen select > the slice you want to use. Continue with the install. When It > completes reboot to single user. > > zfs create -V rpool/dump rpool > zfs create -V rpool/swap rpool > dumpadm -d /dev/zvol/dsk/rpool/dump > swap -a /dev/zvol/dsk/rpool/swap > touch /reconfigure > init 6 > > You should now come up with rpool in the 100 GB slice. > > That said, we can boot from RAIDZ now. The text-install on the > Desktop live image will let you create a mirror, RAIDZ1 or RAIDZ2 and > will take care of all the label stuff. Despite the statement that it > will only use 2 TB, it in fact uses the entire disk. > > It creates a 250 MB s0 slice and the rest of the disk in s1. The 250 > MB slice is labeled "System", but I've not seen any explanation of > it. I've also created RAIDZ2 pools by hand and used F5 to install > into them. F5 appears to be intended to install into a new BE in an > existing pool, hence the need to set up dump and swap by hand. > > Ultimately I decided I didn't care about 1 GB of unused space in 16 > TB of space. So I just went with the text-install created RAIDZ2 > pool. The reconfigure on the first boot after the install is critical > to getting 2021.04.05 up properly. > > Reg > On Saturday, May 1, 2021, 02:57:23 AM CDT, Michelle < > miche...@msknight.com> wrote: > > OK - I appear to be well out of touch. > > I booted the installer and went into prompt. > > Used format (only 1 x 6TB drive in the machine at this point) to > create > a new Solaris 2 partition table and then fdisk'd an all free hog to > partition 1, giving partition 0 100gig. > > I noticed that it must have gone on for 40 odd partitions, and also > there was none of the usual backup and reserved partitions for 2, 8 > and > 9 as I saw before. > > On installation of OI, I selected the drive and got the warning... > "you have chosen a gpt labelled disk. installing onto a gpt labelled > disk will cause the loss of all existing data" > > Out of interest I continued through and got the options for whole > disk > or partition (MBR) ... the second of which gave me a 2Tb Solaris 2 > partition in the list. > > I did try F5 to change partition, but it just took me straight back > to > the installation menu at the start again. > > Things have obviously moved on and I haven't kept pace. > > I now have to work out how to do this on a gpt drive. > > If anyone has any notes, I'd be grateful. > > Michelle. > > > On Sat, 2021-05-01 at 08:31 +0100, Michelle wrote: > > Well, I looked over my notes and the last time I did this was in > > 2014. > > > > My preference has always been to run OI on its own drive and have > > the > > main ZFS tank as a "whole drive" basis. However, thanks to the > > QNAP, > > that's changed. > > > > In 2014 I did a test. I took two 40gig drives and did the > > partitions > > as > > an all free hog on partition 0 ... I was simply testing the ability > > to > > configure rpool on two drives and have both active, so if one > > failed > > the other would keep running the OS. > > > > My immediate thought is to have 100gig for the OS on partition 0 > > and > > the rest on partition 1. Also, turn on auto expand for the tank > > pool > > and off for the rpool. > > > > That's my gut feel. > > > > Anyo
Re: [OpenIndiana-discuss] Partitions for co-exist NAS drive
I'm just going to try breaking it down to two Solaris 2 partitions and see where that takes me. Michelle. On Sat, 2021-05-01 at 08:56 +0100, Michelle wrote: > OK - I appear to be well out of touch. > > I booted the installer and went into prompt. > > Used format (only 1 x 6TB drive in the machine at this point) to > create > a new Solaris 2 partition table and then fdisk'd an all free hog to > partition 1, giving partition 0 100gig. > > I noticed that it must have gone on for 40 odd partitions, and also > there was none of the usual backup and reserved partitions for 2, 8 > and > 9 as I saw before. > > On installation of OI, I selected the drive and got the warning... > "you have chosen a gpt labelled disk. installing onto a gpt labelled > disk will cause the loss of all existing data" > > Out of interest I continued through and got the options for whole > disk > or partition (MBR) ... the second of which gave me a 2Tb Solaris 2 > partition in the list. > > I did try F5 to change partition, but it just took me straight back > to > the installation menu at the start again. > > Things have obviously moved on and I haven't kept pace. > > I now have to work out how to do this on a gpt drive. > > If anyone has any notes, I'd be grateful. > > Michelle. > > > On Sat, 2021-05-01 at 08:31 +0100, Michelle wrote: > > Well, I looked over my notes and the last time I did this was in > > 2014. > > > > My preference has always been to run OI on its own drive and have > > the > > main ZFS tank as a "whole drive" basis. However, thanks to the > > QNAP, > > that's changed. > > > > In 2014 I did a test. I took two 40gig drives and did the > > partitions > > as > > an all free hog on partition 0 ... I was simply testing the ability > > to > > configure rpool on two drives and have both active, so if one > > failed > > the other would keep running the OS. > > > > My immediate thought is to have 100gig for the OS on partition 0 > > and > > the rest on partition 1. Also, turn on auto expand for the tank > > pool > > and off for the rpool. > > > > That's my gut feel. > > > > Anyone got any advice to offer please, before I commit finger to > > keyboard? > > > > Michelle. > > > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Partitions for co-exist NAS drive
OK - I appear to be well out of touch. I booted the installer and went into prompt. Used format (only 1 x 6TB drive in the machine at this point) to create a new Solaris 2 partition table and then fdisk'd an all free hog to partition 1, giving partition 0 100gig. I noticed that it must have gone on for 40 odd partitions, and also there was none of the usual backup and reserved partitions for 2, 8 and 9 as I saw before. On installation of OI, I selected the drive and got the warning... "you have chosen a gpt labelled disk. installing onto a gpt labelled disk will cause the loss of all existing data" Out of interest I continued through and got the options for whole disk or partition (MBR) ... the second of which gave me a 2Tb Solaris 2 partition in the list. I did try F5 to change partition, but it just took me straight back to the installation menu at the start again. Things have obviously moved on and I haven't kept pace. I now have to work out how to do this on a gpt drive. If anyone has any notes, I'd be grateful. Michelle. On Sat, 2021-05-01 at 08:31 +0100, Michelle wrote: > Well, I looked over my notes and the last time I did this was in > 2014. > > My preference has always been to run OI on its own drive and have the > main ZFS tank as a "whole drive" basis. However, thanks to the QNAP, > that's changed. > > In 2014 I did a test. I took two 40gig drives and did the partitions > as > an all free hog on partition 0 ... I was simply testing the ability > to > configure rpool on two drives and have both active, so if one failed > the other would keep running the OS. > > My immediate thought is to have 100gig for the OS on partition 0 and > the rest on partition 1. Also, turn on auto expand for the tank pool > and off for the rpool. > > That's my gut feel. > > Anyone got any advice to offer please, before I commit finger to > keyboard? > > Michelle. > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Partitions for co-exist NAS drive
Well, I looked over my notes and the last time I did this was in 2014. My preference has always been to run OI on its own drive and have the main ZFS tank as a "whole drive" basis. However, thanks to the QNAP, that's changed. In 2014 I did a test. I took two 40gig drives and did the partitions as an all free hog on partition 0 ... I was simply testing the ability to configure rpool on two drives and have both active, so if one failed the other would keep running the OS. My immediate thought is to have 100gig for the OS on partition 0 and the rest on partition 1. Also, turn on auto expand for the tank pool and off for the rpool. That's my gut feel. Anyone got any advice to offer please, before I commit finger to keyboard? Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] OI on USB flash drive
Hi Reg, Thanks for the efforts. I did try the GUI installer but got the same problem at initial boot. My issue does seem to be the USB controller. It is looking like I'll have to do the slice technique for the rpool on the data drives. They're a pair of 6tb wd reds so I won't loose a lot :-) ... but they're getting a little old and I didn't want to risk the OS on them at well... but it is what it is. I last did it several years ago so I'll have to look it up again! The 8tb I took out of the QNAP and put in an N54L, zfs quickly flagged one of them (which is now RMA) whereas the QNAP was still happy to use it... so I do prefer ZFS for the error checking. Till mentioned SmartOS, so I'll take a look at that, but I by far prefer ZFS wherever I can get it. Many thanks for your efforts and write up. Michelle. On Fri, 2021-04-30 at 22:55 +, Reginald Beardsley via openindiana- discuss wrote: > > > While I don't have prior experience booting from flash drives, I > agree with Till's recommendation to try a USB hard drive. > > Have you tried installing to a regular disk drive and booting from > that? If you can I recommend doing that and skipping the flash drive > boot. The RAIDZ bootloader seems to work though I don't have the 10 > years of experience I have with using 2 slices to provide a mirrored > rpool and a RAIDZ export pool. > > Reg > On Friday, April 30, 2021, 05:22:03 PM CDT, Till Wegmueller < > toaster...@gmail.com> wrote: > > Hi Michelle > > I don't know how happy you are going to get with a QNAP Hardware as > base > for OI. QNAP is very exotic Hardware. It may be USB what they are > using > in princiapl, but unless you are sure the Controllers are actually > the > Controllers they are saying they are and the Storage is not a SD card > in > Hiding, you are going to have a bad time. ZFS does not assume the > Storage to be uncorruptable, and thus far every SD-Card I found was > so > terrible, that one meeting with ZFS showed just how much dataloss > these > cards have. > > QNAP and Synology usually use RAMDISK based approaches which run > fully > in RAM and only use the Storage to boot. > > SmartOS does the same, so this could work. > > In my career and with friends I have seen many OS installation to > USB > drives and SD-cards go horribly wrong. Those storage Options usually > don't survive for long and then the whole OS is gone. I've seen it > with > FreeNAS(FreeBSD), Linux and illumos based Oses. I use Disks or SSD's > as > Storage for the Pools. The Cable can be USB. But it needs to be a > Disk > and not a flashdrive to at least Guarantee some use. > > My Hardware recommendations for Cheap Home Servers for NAS on a > budget > include HP Microservers Gen8 (Gen10 works too, but they spew warnings > on > boot) SuperMicro Mini-ITX Boards. HP Workstations, or Workstations > and > Gaming and Office Rigs in general and of course Dan McDonalds HDC3.0 > [0] > > -Till > > [0] https://kebe.com/blog/?tag=HDC > > On 30.04.21 17:19, Michelle wrote: > > Good suggestions, but sadly didn't work. > > > > Using option 2 single user also fails with the same errors. > > > > I've also tried the USB interface in various booting modes, > > unfortunately none of which work. > > > > I took a look at the variables in the boot options, just for the > > sake > > of it, and couldn't find anything obvious that would help > > > > Michelle. > > > > On Fri, 2021-04-30 at 20:00 +, Reginald Beardsley via > > openindiana- > > discuss wrote: > > > Try selecting the "reconfigure" option when you first boot the > > > installed image. > > > > > > It is *supposed* to reconfigure, but I spent 50 hrs before I > > > found > > > out how to get 2021.04.05 to boot after the install. I thought > > > the > > > issue was the particular nVIDIA driver, but that was not the > > > case. > > > > > > I also got a fresh install of 2021.04.05 to boot properly by > > > coming > > > up single user, "touch /reconfigure; init 6". > > > > > > Good luck, > > > Reg > > > > > > > > > On Friday, April 30, 2021, 02:50:50 PM CDT, Michelle < > > > miche...@msknight.com> wrote: > > > > > > OK, this is where things are going to get a bit awkward. > > > > > > Scenario – QNAP NAS TS-251 decided to install their own Malware > > > Remover > > > which tells me it’s removed files
Re: [OpenIndiana-discuss] OI on USB flash drive
Hi Till, Many thanks for the detailed response. My target is an enterprise Sandisk SSD on the end of a Sata/USB controller. I think I can get it in the case. I've tried a few combinations and have one that boots, so I'm fairly confident that this is now down to tweeking; but I might have reached the end of the road with the USB approach because, as you say, they are a bit exotic. I do have a couple of HP N54L microservers in service for about five years, with the custom BIOS flash so I have hot plug/play. The newer microservers still run to a grand with a decent spec, which is a bit rich for me at the moment; I need to save. This is now just a case of not letting the QNAP go to landfill. You're correct... my model is a half gig flash memory module on a USB header and it unpacks itself into memory. With that gone, I'm at the 2015 American Megatrends BIOS controlling the boot, but I haven't found an update that I trust for that motherboard yet. Thanks for the tip on SmartOS - I'll look it up. Michelle. On Fri, 2021-04-30 at 19:21 -0300, Till Wegmueller wrote: > Hi Michelle > > I don't know how happy you are going to get with a QNAP Hardware as > base > for OI. QNAP is very exotic Hardware. It may be USB what they are > using > in princiapl, but unless you are sure the Controllers are actually > the > Controllers they are saying they are and the Storage is not a SD card > in > Hiding, you are going to have a bad time. ZFS does not assume the > Storage to be uncorruptable, and thus far every SD-Card I found was > so > terrible, that one meeting with ZFS showed just how much dataloss > these > cards have. > > QNAP and Synology usually use RAMDISK based approaches which run > fully > in RAM and only use the Storage to boot. > > SmartOS does the same, so this could work. > > In my career and with friends I have seen many OS installation to > USB > drives and SD-cards go horribly wrong. Those storage Options usually > don't survive for long and then the whole OS is gone. I've seen it > with > FreeNAS(FreeBSD), Linux and illumos based Oses. I use Disks or SSD's > as > Storage for the Pools. The Cable can be USB. But it needs to be a > Disk > and not a flashdrive to at least Guarantee some use. > > My Hardware recommendations for Cheap Home Servers for NAS on a > budget > include HP Microservers Gen8 (Gen10 works too, but they spew warnings > on > boot) SuperMicro Mini-ITX Boards. HP Workstations, or Workstations > and > Gaming and Office Rigs in general and of course Dan McDonalds HDC3.0 > [0] > > -Till > > [0] https://kebe.com/blog/?tag=HDC > > On 30.04.21 17:19, Michelle wrote: > > Good suggestions, but sadly didn't work. > > > > Using option 2 single user also fails with the same errors. > > > > I've also tried the USB interface in various booting modes, > > unfortunately none of which work. > > > > I took a look at the variables in the boot options, just for the > > sake > > of it, and couldn't find anything obvious that would help > > > > Michelle. > > > > On Fri, 2021-04-30 at 20:00 +, Reginald Beardsley via > > openindiana- > > discuss wrote: > > > Try selecting the "reconfigure" option when you first boot the > > > installed image. > > > > > > It is *supposed* to reconfigure, but I spent 50 hrs before I > > > found > > > out how to get 2021.04.05 to boot after the install. I thought > > > the > > > issue was the particular nVIDIA driver, but that was not the > > > case. > > > > > > I also got a fresh install of 2021.04.05 to boot properly by > > > coming > > > up single user, "touch /reconfigure; init 6". > > > > > > Good luck, > > > Reg > > > > > > > > > On Friday, April 30, 2021, 02:50:50 PM CDT, Michelle < > > > miche...@msknight.com> wrote: > > > > > > OK, this is where things are going to get a bit awkward. > > > > > > Scenario – QNAP NAS TS-251 decided to install their own Malware > > > Remover > > > which tells me it’s removed files, but doesn’t tell me which ones > > > it’s > > > removed, and I can’t kill the remover itself. If I Ssh in and > > > remove > > > it, it simply re-installs itself. I'm not a fan of malware > > > removers > > > that don't tell you what files they've removed. > > > > > > So… the decision to rip it apart and install my own operating > > > system. > > > Aft
Re: [OpenIndiana-discuss] OI on USB flash drive
Good suggestions, but sadly didn't work. Using option 2 single user also fails with the same errors. I've also tried the USB interface in various booting modes, unfortunately none of which work. I took a look at the variables in the boot options, just for the sake of it, and couldn't find anything obvious that would help Michelle. On Fri, 2021-04-30 at 20:00 +, Reginald Beardsley via openindiana- discuss wrote: > Try selecting the "reconfigure" option when you first boot the > installed image. > > It is *supposed* to reconfigure, but I spent 50 hrs before I found > out how to get 2021.04.05 to boot after the install. I thought the > issue was the particular nVIDIA driver, but that was not the case. > > I also got a fresh install of 2021.04.05 to boot properly by coming > up single user, "touch /reconfigure; init 6". > > Good luck, > Reg > > > On Friday, April 30, 2021, 02:50:50 PM CDT, Michelle < > miche...@msknight.com> wrote: > > OK, this is where things are going to get a bit awkward. > > Scenario – QNAP NAS TS-251 decided to install their own Malware > Remover > which tells me it’s removed files, but doesn’t tell me which ones > it’s > removed, and I can’t kill the remover itself. If I Ssh in and remove > it, it simply re-installs itself. I'm not a fan of malware removers > that don't tell you what files they've removed. > > So… the decision to rip it apart and install my own operating system. > After all, it’s on an American Megatrends bios. > > The 2gig of RAM was replaced with 8, and the storage is a half gig > USB > flash module. So the flash module has to be replaced also. > > I thus have a USB header unit and I’ve also tried to install Hipster > to > a flash drive connected via a USB harness, and also to a USB key > directly. > > On install, I get an error… > > openindiana drm: WARNING: [drm:pci_dev_create:93] ddi_prop_get_int() > failed > > … which repeats three or four times during the installation. > > After installation is finished, the installation OS goes down, but > says… > > openindiana genunix:WARNING: xhci has no quiesce() > reboot: not all drivers have implemented quiesce(9E) > Please see veradm/messages for drivers that haven’t implemented > quiesce(9E) > Failed to process boot arguments from Boot Environment. > Falling back to regular reboot > > Then on reboot, the boot comes up but gives… > > WARNING: pci@0,0pci8086,f35@14/storage@3/disk@0,0 (sd0): > Commend failed to complete… Device is gone > Warning: Pool “rpool” has encountered an uncorrectable I/O failure > and > has been suspended; “zpool clear” will be required before the poor > can > be written to. > Warning: pci@0,0/pci8086/f35@14/storage@3 (scsa2usb0): Reinserted > device is accessible again. > Warning: Pool “rpool” has encountered an incorrectable I/O failure > and > has been suspended; “zpool clear” will be required before the poor > can > be written to. > > ...that last repeats three times and the unit sits there. > > I did try searching for devices with no drivers, but the only thing > that came up were the two processor cores. > > I did try creating a Solaris 2 partition and installing to a > partition > instead of the EFI whole disk option, but that failed on fdisk and > didn’t even install the files. I had to destroy the partition table > in > order to get back to getting the installation to work. > > Also, in the installer, when I tried to go to a partition > installation > and pressed F5 to change the partition type, it would go back to the > installation menu again, so I’m presuming that it just doesn’t like > the > USB as a target. > > I’ve got an ugly feeling that I’ve got no way of getting around > this. > Yes, I know I can put the SATA drives in and make a slice for rpool, > but where’s the fun in that? > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] OI on USB flash drive
OK, this is where things are going to get a bit awkward. Scenario – QNAP NAS TS-251 decided to install their own Malware Remover which tells me it’s removed files, but doesn’t tell me which ones it’s removed, and I can’t kill the remover itself. If I Ssh in and remove it, it simply re-installs itself. I'm not a fan of malware removers that don't tell you what files they've removed. So… the decision to rip it apart and install my own operating system. After all, it’s on an American Megatrends bios. The 2gig of RAM was replaced with 8, and the storage is a half gig USB flash module. So the flash module has to be replaced also. I thus have a USB header unit and I’ve also tried to install Hipster to a flash drive connected via a USB harness, and also to a USB key directly. On install, I get an error… openindiana drm: WARNING: [drm:pci_dev_create:93] ddi_prop_get_int() failed … which repeats three or four times during the installation. After installation is finished, the installation OS goes down, but says… openindiana genunix:WARNING: xhci has no quiesce() reboot: not all drivers have implemented quiesce(9E) Please see veradm/messages for drivers that haven’t implemented quiesce(9E) Failed to process boot arguments from Boot Environment. Falling back to regular reboot Then on reboot, the boot comes up but gives… WARNING: pci@0,0pci8086,f35@14/storage@3/disk@0,0 (sd0): Commend failed to complete… Device is gone Warning: Pool “rpool” has encountered an uncorrectable I/O failure and has been suspended; “zpool clear” will be required before the poor can be written to. Warning: pci@0,0/pci8086/f35@14/storage@3 (scsa2usb0): Reinserted device is accessible again. Warning: Pool “rpool” has encountered an incorrectable I/O failure and has been suspended; “zpool clear” will be required before the poor can be written to. ...that last repeats three times and the unit sits there. I did try searching for devices with no drivers, but the only thing that came up were the two processor cores. I did try creating a Solaris 2 partition and installing to a partition instead of the EFI whole disk option, but that failed on fdisk and didn’t even install the files. I had to destroy the partition table in order to get back to getting the installation to work. Also, in the installer, when I tried to go to a partition installation and pressed F5 to change the partition type, it would go back to the installation menu again, so I’m presuming that it just doesn’t like the USB as a target. I’ve got an ugly feeling that I’ve got no way of getting around this. Yes, I know I can put the SATA drives in and make a slice for rpool, but where’s the fun in that? ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] ZFS SMB unauthenticated user on share
I have a ZFS SMB share, let's say, "public" and mounted at /mnt/public with permissions of 744. Fresh install of Hipster. I would like to be able for someone to mount it read-only without being prompted for user name and password. I thought it was... zfs set share.smb.guestok=on public ...but apparently this isn't it and I can't find what the property is. I have tried a few variations. I did find this... https://docs.oracle.com/cd/E26502_01/html/E29004/managingsmbshares.html#enableguestaccesstask ...but that's not working either. Grateful for any help. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Graphics card CLI enquiry
It's just text only. The N54L only has a VGA connector and I need a DVI for connection to my KVM. The KVM won't take analogue over DVI so I need a half height card to do the job. Michelle On Tue, 2021-04-27 at 08:20 +0200, Stephan Althaus wrote: > On 04/26/21 03:21 PM, Michelle wrote: > > Many thanks... > > > > prtdiag -v told me > > > > Upgradeable Slots > > > > ID StatusType Description > > --- - > > 1 in usePCI Express PCIE1-J5, Advanced Micro Devices, > > Inc. > > [AMD/ATI] RV370 [Radeon X600/X600 SE] (vgatext) > > 2 available PCI Express PCIE2-J6 > > > > > > On Mon, 2021-04-26 at 14:39 +0200, Stephan Althaus wrote: > > > On 04/26/21 02:28 PM, Michelle wrote: > > > > As QNAP are enforcing their malware removal tool out, which is > > > > then > > > > deleting files but has no log of which files it's deleted (and > > > > you > > > > can't turn the malware removal tool off... it just re-installs > > > > itself > > > > and keeps running) > > > > > > > > ...I'm having to press my backup N54L into use and install OI > > > > on > > > > it. > > > > > > > > The problem is, I need to query my currently running N54L with > > > > OI > > > > to > > > > find out what graphics card is in it, so I can order a second > > > > one. > > > > > > > > The usual linux command isn't there so I was wondering what CLI > > > > command > > > > I run to find out what's in there please, as it isn't exactly a > > > > command > > > > I use every day :-) > > > > > > > > > > > > ___ > > > > openindiana-discuss mailing list > > > > openindiana-discuss@openindiana.org > > > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > > Hello! > > > > > > prtconf -d -D > > > > > > is your friend. if you need a lot of detail, try > > > > > > prtconf -v > > > > > > > > > hth > > > > > > Stephan > > > > > > > > > ___ > > > openindiana-discuss mailing list > > > openindiana-discuss@openindiana.org > > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > Hi! > > OI works best with NVidia quadro cards, support for amd is very > limited > from what i know. > > Stephan > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Graphics card CLI enquiry
Many thanks... prtdiag -v told me Upgradeable Slots ID StatusType Description --- - 1 in usePCI Express PCIE1-J5, Advanced Micro Devices, Inc. [AMD/ATI] RV370 [Radeon X600/X600 SE] (vgatext) 2 available PCI Express PCIE2-J6 On Mon, 2021-04-26 at 14:39 +0200, Stephan Althaus wrote: > On 04/26/21 02:28 PM, Michelle wrote: > > As QNAP are enforcing their malware removal tool out, which is then > > deleting files but has no log of which files it's deleted (and you > > can't turn the malware removal tool off... it just re-installs > > itself > > and keeps running) > > > > ...I'm having to press my backup N54L into use and install OI on > > it. > > > > The problem is, I need to query my currently running N54L with OI > > to > > find out what graphics card is in it, so I can order a second one. > > > > The usual linux command isn't there so I was wondering what CLI > > command > > I run to find out what's in there please, as it isn't exactly a > > command > > I use every day :-) > > > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > https://openindiana.org/mailman/listinfo/openindiana-discuss > > Hello! > > prtconf -d -D > > is your friend. if you need a lot of detail, try > > prtconf -v > > > hth > > Stephan > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > https://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Graphics card CLI enquiry
As QNAP are enforcing their malware removal tool out, which is then deleting files but has no log of which files it's deleted (and you can't turn the malware removal tool off... it just re-installs itself and keeps running) ...I'm having to press my backup N54L into use and install OI on it. The problem is, I need to query my currently running N54L with OI to find out what graphics card is in it, so I can order a second one. The usual linux command isn't there so I was wondering what CLI command I run to find out what's in there please, as it isn't exactly a command I use every day :-) ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ACL problem
Thanks Gea, Unfortunately I'm building this server for a friend. I only have Linux Mint machines, and the OI box. I used to use Netware, and when I started with AD, I had a bit of a freak out, as it was sort of completely the reverse! Your notes are valuable. I'll put them in my keep folder. For this case, after what Peter said, the user/group name is the most important thing, so I'll tell my friend just to not do any file transfers at 3am :-D Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ACL problem
On Sun, 2015-12-13 at 17:11 +, Peter Tribble wrote: > Unfortunately, that's something you can't do. Ah. Bugger. Looks like back to cron jobs and I'll abandon the ACL approach then! Many thanks for taking the time to respond. Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ACL problem
I'm definitely having problems with this inherit_only flag. What starts as this... drwxrwxrwx+ 2 Joe Family 2 Dec 13 15:40 guest_folder 0:owner@:list_directory/read_data/add_file/write_data/add_subdirectory /append_data/read_xattr/execute/delete_child/read_attributes /delete/read_acl:file_inherit/dir_inherit:allow 1:group@:list_directory/read_data/add_file/write_data/add_subdirectory /append_data/read_xattr/execute/delete_child/read_attributes /delete/read_acl:file_inherit/dir_inherit:allow 2:everyone@:list_directory/read_data/add_file/write_data /add_subdirectory/append_data/read_xattr/execute/delete_child /read_attributes/delete/read_acl:file_inherit/dir_inherit:allow ... if a guest (who has access under "everyone) writes a file or makes a directory in that folder, then the ownership is not Joe, but the guest account. Adding the inherit_only flag changes to this... d-+ 2 Joe Family 2 Dec 13 15:40 guest_folder 0:owner@:list_directory/read_data/add_file/write_data/add_subdirectory /append_data/read_xattr/execute/delete_child/read_attributes /delete/read_acl:file_inherit/dir_inherit/inherit_only:allow 1:group@:list_directory/read_data/add_file/write_data/add_subdirectory /append_data/read_xattr/execute/delete_child/read_attributes /delete/read_acl:file_inherit/dir_inherit/inherit_only:allow 2:everyone@:list_directory/read_data/add_file/write_data /add_subdirectory/append_data/read_xattr/execute/delete_child /read_attributes/delete/read_acl:file_inherit/dir_inherit /inherit_only:allow ...and not even Joe can see the directory guest_folder in an SFTP listing, despite being the owner and having the rights. There's something obvious going on here that I clearly haven't got to grips with. All I'm trying to do, is ensure that all files written within a directory, have the same ownership as the directory itself, no matter what account actually writes them. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ACL problem
I think part of my problem may have been the "inherit_only" flag which is described as, "Inherit the ACL from the parent directory but applies only to newly created files or subdirectories and not the directory itself. This flag requires the file_inherit flag, the dir_inherit flag or both to indicate what to inherit" ...but the moment I applied that to a directory, it immediately took the permissions from the parent directory of the parent, whereas I was only expecting it to apply to other files and sub-directories that were created below it, later on. I might be making progress in my understanding of where I'm going wrong, but it is only one step forward. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] ACL problem
Hi Folks, I've been having problems with ACL's. My understanding is just not getting there and reading the Solaris ZFS Administration Guide isn't sinking in. This is roughly what I want to do... User, Joe Group Family Other Set up a directory, (say Movies) to which Joe is the owner. Family can access and read, but not write, and Other can't access at all. Joe's ownership of files and all ACL permissions are to be propagated to any file or directory created, by whatever means, be it SFTP or CIFS access to the share. There is another situation where I want "Family" to be able to create, delete and modify files and directories freely, but for ownership of file and directories to still explicitly be Joe. There is a chance that my previous experiments may be messing things up, so the other question is, does deleting a directory and re-instating it clear all previous extended permission settings, or must I issue a command to clear all flags before my next failed attempt :-) ... as obviously deleting the .$EXTEND file is not allowed. Many thanks, Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Persistent permissions
Reg, A good idea. The long story short is that this is a home system on a N40L server. Simple set up, three drives in a RaidZ. 1) There is a "guest" area where guests can transfer data to and from the server (the owner can copy data there specifically for the guest without having to give them access to the main structure) ... and as such the guests will need to be able to read and write to this area. ...the issue here is that any files that the user copies to that area will have their permissions and ownership, so there is a need for those files to then be changed so that the guest can get at them ...also there is an issue that any files moved from the guest area in to the main storage will have the guests credentials and need to be changed appropriately according to which area they have been moved to. 2) A read-write area where the main user can use the files with no issue. 3) A series of read-only folders where a different, elevated user, is logged on deliberately in order to manipulate files in the read only areas. This is so that any virus that gets on to a client, can not damage these areas as the elevated user is not in common use. Also, it helps prevent accidental deletion. ...so files which are copied by the elevated user have to also be accessible by the read-write user, and also any files that the read-write user puts out of read-write and in to read-only will also require a permissions and ownership change so that they are put out of reach of potential change. And in all this, the files can be moved/changed in the system via either SMB (which the guests and read-only users are going to use) and SFTP (which is most likely the vector that he will go in to the server for the read-write elevated user ops.) Michelle. On Wed, 20 Aug 2014 13:49:07 -0700 Reginald Beardsley via openindiana-discuss wrote: > Michelle, > > I would like to suggest articulating the policy you want to > enforce. You started with a proposed solution that you were having > trouble getting to work satisfactorily. > > I had the strong impression that a simpler approach would do what you > want, but no information upon which to suggest a solution. All I > knew for certain was that it all sounded rather complicated. > > Trying to implement a policy you can't clearly state is not likely to > succeed no matter how hard you work at it. At best you get something > that seems to work until you discover it had unintended consequences > which are now causing problems. > > Reg > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > http://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Persistent permissions
Thanks, I'll try this tomorrow. Just off to sleep and my forehead is bleeding! I'm not using AD, just everything pure and local! Michelle On Wed, 20 Aug 2014 15:48:33 -0400 (EDT) "w...@vandenberge.us" wrote: > Michelle, > > Set the file-system to inherit ACL's and then set the permissions > correctly. Something like this: > > zfs set aclinherit=passthrough pool01/fs01 > /usr/bin/chmod > A=group:writegroup:rwxpdDaARWcCos:fd-:allow,group:readgroup:r-x---a-R-c--s:fd-:allow > /pool01/fs01 > > If you already have data in the fs add -R to chmod. > > This is of course assuming you mounted pool01/fs01 on /pool01/fs01 > and that your AD group to unix group mapping is set up correctly > (idmap) if you don't use AD to resolve groups. > > > W > > > > > > > > > On August 20, 2014 at 3:02 PM Michelle Knight > > wrote: > > > > > > Hi John, > > > > Thanks. I'm looking through it now. > > > > It is a comprehensive article with not many examples for the newbie > > and because of that, I'm having difficulty decoding it, but if I > > continue looking hard enough, I'll get it! > > > > From what I can work out, I need to set the trivial ACLs because the > > non-trivial ones aren't propagated. > > > > I'll get there! > > > > Michelle > > > > > > On Wed, 20 Aug 2014 14:51:43 -0400 > > John D Groenveld wrote: > > > > > In message <20140820194500.285b91cf@fitpc3>, Michelle Knight > > > writes: > > > >It has it, but I get... > > > > > > > >"File system doesn't support aclent_t style ACL's." > > > > > > > >..so I'm obviously doing something wrong. > > > > > > See chmod(1) > > > http://docs.oracle.com/cd/E19120-01/open.solaris/817-2271/ftyxi/index.html > > > > > > > > > > John > > > groenv...@acm.org > > > > > > > > > ___ > > > openindiana-discuss mailing list > > > openindiana-discuss@openindiana.org > > > http://openindiana.org/mailman/listinfo/openindiana-discuss > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > http://openindiana.org/mailman/listinfo/openindiana-discuss > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > http://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Persistent permissions
OK... I execute ... chmod A0=owner@:list_directory/read_data/add_file/write_data/add_subdirectory/append_data/read_xattr/write_xattr/execute/read_attributes/write_attributes/read_acl/write_acl/write_owner/synchronize:file_inherit:allow guest chmod A1=group@:list_directory/read_data/add_file/write_data/add_subdirectory/append_data/read_xattr/write_xattr/execute/read_attributes/write_attributes/read_acl/write_acl/write_owner/synchronize:file_inherit:allow guest chmod A2=everyone@:list_directory/read_data/add_file/write_data/add_subdirectory/append_data/read_xattr/write_xattr/execute/read_attributes/write_attributes/read_acl/write_acl/write_owner/synchronize:file_inherit:deny guest ...which gives... Janitor@mother:/mnt/stuff# ls -dv guest drwxrwx---+ 2 myguest guest 3 Aug 20 20:26 guest 0:owner@:list_directory/read_data/add_file/write_data/add_subdirectory /append_data/read_xattr/write_xattr/execute/read_attributes /write_attributes/read_acl/write_acl/write_owner/synchronize :file_inherit:allow 1:group@:list_directory/read_data/add_file/write_data/add_subdirectory /append_data/read_xattr/write_xattr/execute/read_attributes /write_attributes/read_acl/write_acl/write_owner/synchronize :file_inherit:allow 2:everyone@:list_directory/read_data/add_file/write_data /add_subdirectory/append_data/read_xattr/write_xattr/execute /read_attributes/write_attributes/read_acl/write_acl/write_owner /synchronize:file_inherit:deny So ... I go in to "guest" and execute "touch test" and get... Janitor@mother:/mnt/stuff/guest# ls -dv test -rw-r--r-- 1 root root 0 Aug 20 20:29 test 0:owner@:read_data/write_data/append_data/read_xattr/write_xattr /read_attributes/write_attributes/read_acl/write_acl/write_owner /synchronize:allow 1:group@:read_data/read_xattr/read_attributes/read_acl/synchronize:allow 2:everyone@:read_data/read_xattr/read_attributes/read_acl/synchronize :allow Not what I was expecting. Neither the ownership or the permissions have propagated. What have I missed please? Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Persistent permissions
Hi John, Thanks. I'm looking through it now. It is a comprehensive article with not many examples for the newbie and because of that, I'm having difficulty decoding it, but if I continue looking hard enough, I'll get it! >From what I can work out, I need to set the trivial ACLs because the non-trivial ones aren't propagated. I'll get there! Michelle On Wed, 20 Aug 2014 14:51:43 -0400 John D Groenveld wrote: > In message <20140820194500.285b91cf@fitpc3>, Michelle Knight writes: > >It has it, but I get... > > > >"File system doesn't support aclent_t style ACL's." > > > >..so I'm obviously doing something wrong. > > See chmod(1) > http://docs.oracle.com/cd/E19120-01/open.solaris/817-2271/ftyxi/index.html > > > > John > groenv...@acm.org > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > http://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Persistent permissions
Sorry, should have included an example command... setfacl -m g:myguest:rwx,d:g:guest:rwx guest On Wed, 20 Aug 2014 12:32:27 -0600 "Dormition Skete (Hotmail)" wrote: > Does OpenIndiana have facl? If it does, “setfacl" should do what you > are looking for. > > http://www.computerhope.com/unix/usetfacl.htm > > > > > On Aug 20, 2014, at 12:14 PM, Michelle Knight > wrote: > > > Hi Folks, > > > > I'm not sure how to do this under ZFS, but here goes... > > > > I have a ZFS share/pool which is accessed via SMB and also SFTP. A > > number of different users can put (and move) files between > > directories and even create directories in some cases. > > > > I would like any file which is uploaded or moved (or new directory > > created) ... to obtain the permissions and levels of the directory > > that it is moved/copied to, regardless of what it had before it was > > moved, and regardless of the account or permissions of the account > > that did the operation, whether it be done via SMB or SFTP. Also, > > any directory then created will have the same influence on any > > files created/copied in to it. > > > > At the moment, I am using a script which will go through the > > directory structures and forcefully re-write the ownerships and > > permissions of every file, but this obviously has an impact as if > > there is a copy operation going on when the script hits that > > particular area, it can re-write the permissions on a file and halt > > the copy process because of a permissions error. This technique > > also has the problem that some users have to wait for the script to > > run, because they won't have access until the file/directory > > permissions have been re-written. > > > > Is there any more elegant way of doing this please? > > > > Michelle. > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > http://openindiana.org/mailman/listinfo/openindiana-discuss > > > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > http://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Persistent permissions
It has it, but I get... "File system doesn't support aclent_t style ACL's." ..so I'm obviously doing something wrong. On Wed, 20 Aug 2014 12:32:27 -0600 "Dormition Skete (Hotmail)" wrote: > Does OpenIndiana have facl? If it does, “setfacl" should do what you > are looking for. > > http://www.computerhope.com/unix/usetfacl.htm > > > > > On Aug 20, 2014, at 12:14 PM, Michelle Knight > wrote: > > > Hi Folks, > > > > I'm not sure how to do this under ZFS, but here goes... > > > > I have a ZFS share/pool which is accessed via SMB and also SFTP. A > > number of different users can put (and move) files between > > directories and even create directories in some cases. > > > > I would like any file which is uploaded or moved (or new directory > > created) ... to obtain the permissions and levels of the directory > > that it is moved/copied to, regardless of what it had before it was > > moved, and regardless of the account or permissions of the account > > that did the operation, whether it be done via SMB or SFTP. Also, > > any directory then created will have the same influence on any > > files created/copied in to it. > > > > At the moment, I am using a script which will go through the > > directory structures and forcefully re-write the ownerships and > > permissions of every file, but this obviously has an impact as if > > there is a copy operation going on when the script hits that > > particular area, it can re-write the permissions on a file and halt > > the copy process because of a permissions error. This technique > > also has the problem that some users have to wait for the script to > > run, because they won't have access until the file/directory > > permissions have been re-written. > > > > Is there any more elegant way of doing this please? > > > > Michelle. > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > http://openindiana.org/mailman/listinfo/openindiana-discuss > > > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > http://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] Persistent permissions
Hi Folks, I'm not sure how to do this under ZFS, but here goes... I have a ZFS share/pool which is accessed via SMB and also SFTP. A number of different users can put (and move) files between directories and even create directories in some cases. I would like any file which is uploaded or moved (or new directory created) ... to obtain the permissions and levels of the directory that it is moved/copied to, regardless of what it had before it was moved, and regardless of the account or permissions of the account that did the operation, whether it be done via SMB or SFTP. Also, any directory then created will have the same influence on any files created/copied in to it. At the moment, I am using a script which will go through the directory structures and forcefully re-write the ownerships and permissions of every file, but this obviously has an impact as if there is a copy operation going on when the script hits that particular area, it can re-write the permissions on a file and halt the copy process because of a permissions error. This technique also has the problem that some users have to wait for the script to run, because they won't have access until the file/directory permissions have been re-written. Is there any more elegant way of doing this please? Michelle. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Static IP problem
Thanks Carsten, That appears to work and survive reboots. An obvious error with the ... ipadm create-if beg0 ...when the interface already exists and is up, but ... job done! Many thanks, Michelle. On Thu, 10 Jul 2014 08:05:00 +0200 real-men-dont-cl...@gmx.net wrote: > forget about ifconfig. It's outdated. In my experience it nowadays is > a never ending cause of trouble. > > That's what I found out as known to work: > > http://wiki.system-admin.info/doku.php?id=openindiana_tips_and_tricks#network > > Hope this may help > > > Carsten > > > -Original message- > From: Michelle Knight > Sent: Thu 10-07-2014 07:55 > Subject: Re: [OpenIndiana-discuss] Static IP problem > To: openindiana-discuss@openindiana.org; > > Incidentally, in case you're wondering, I'm rebuilding the box > > between tests ... to ensure that what I'm doing, works right. > > > > On Thu, 10 Jul 2014 07:43:56 +0200 > > "Udo Grabowski (IMK)" wrote: > > > > > On 10/07/2014 07:34, Michelle Knight wrote: > > > > OK, this is what I'm doing now ... > > > > > > > > svcadm disable svc:/network/physical:nwam > > > > cp /etc/nsswitch.dns /etc/nsswitch.conf > > > > ifconfig bge0 plumb > > > > ifconfig bge0 192.168.0.25 netmask 255.255.255.0 up > > > > route -p add default 192.168.0.1 > > > > > > > > ...but it doesn't survive a reboot and I can't see anyone using > > > > a persistent option anywhere. > > > > > > You forgot to 'svcadm enable physical:default'. > > > Also, put the default router into /etc/defaultrouter, > > > and put the host names into /etc/hosts and add the > > > /etc/hostname.bge0 file with the corresponding host > > > name (or IP) inside the file. > > > > > > > > > ___ > > > openindiana-discuss mailing list > > > openindiana-discuss@openindiana.org > > > http://openindiana.org/mailman/listinfo/openindiana-discuss > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > http://openindiana.org/mailman/listinfo/openindiana-discuss > > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > http://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Static IP problem
Thanks Carsten, I'll rebuild the box again and try this. Michelle. On Thu, 10 Jul 2014 08:05:00 +0200 real-men-dont-cl...@gmx.net wrote: > forget about ifconfig. It's outdated. In my experience it nowadays is > a never ending cause of trouble. > > That's what I found out as known to work: > > http://wiki.system-admin.info/doku.php?id=openindiana_tips_and_tricks#network > > Hope this may help > > > Carsten > > > -Original message- > From: Michelle Knight > Sent: Thu 10-07-2014 07:55 > Subject: Re: [OpenIndiana-discuss] Static IP problem > To: openindiana-discuss@openindiana.org; > > Incidentally, in case you're wondering, I'm rebuilding the box > > between tests ... to ensure that what I'm doing, works right. > > > > On Thu, 10 Jul 2014 07:43:56 +0200 > > "Udo Grabowski (IMK)" wrote: > > > > > On 10/07/2014 07:34, Michelle Knight wrote: > > > > OK, this is what I'm doing now ... > > > > > > > > svcadm disable svc:/network/physical:nwam > > > > cp /etc/nsswitch.dns /etc/nsswitch.conf > > > > ifconfig bge0 plumb > > > > ifconfig bge0 192.168.0.25 netmask 255.255.255.0 up > > > > route -p add default 192.168.0.1 > > > > > > > > ...but it doesn't survive a reboot and I can't see anyone using > > > > a persistent option anywhere. > > > > > > You forgot to 'svcadm enable physical:default'. > > > Also, put the default router into /etc/defaultrouter, > > > and put the host names into /etc/hosts and add the > > > /etc/hostname.bge0 file with the corresponding host > > > name (or IP) inside the file. > > > > > > > > > ___ > > > openindiana-discuss mailing list > > > openindiana-discuss@openindiana.org > > > http://openindiana.org/mailman/listinfo/openindiana-discuss > > > > ___ > > openindiana-discuss mailing list > > openindiana-discuss@openindiana.org > > http://openindiana.org/mailman/listinfo/openindiana-discuss > > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > http://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Static IP problem
That didn't survive a reboot either. And the ipadm command comes back with the same error as before, "could not create address: Persistent operation on temporary object" I must have got something really wrong here. On Thu, 10 Jul 2014 07:43:56 +0200 "Udo Grabowski (IMK)" wrote: > On 10/07/2014 07:34, Michelle Knight wrote: > > OK, this is what I'm doing now ... > > > > svcadm disable svc:/network/physical:nwam > > cp /etc/nsswitch.dns /etc/nsswitch.conf > > ifconfig bge0 plumb > > ifconfig bge0 192.168.0.25 netmask 255.255.255.0 up > > route -p add default 192.168.0.1 > > > > ...but it doesn't survive a reboot and I can't see anyone using a > > persistent option anywhere. > > You forgot to 'svcadm enable physical:default'. > Also, put the default router into /etc/defaultrouter, > and put the host names into /etc/hosts and add the > /etc/hostname.bge0 file with the corresponding host > name (or IP) inside the file. > > > ___ > openindiana-discuss mailing list > openindiana-discuss@openindiana.org > http://openindiana.org/mailman/listinfo/openindiana-discuss ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss