Re: [OpenIndiana-discuss] OI Hipster becomes unreachable over network after a certain length of uptime
On 4/11/2022 9:14 AM, Bob Friesenhahn wrote: On Sun, 10 Apr 2022, Judah Richardson wrote: OK. The Solaris documentation I linked to says that Solaris (and presumably distros downstream of that codebase) expects the DHCP server to be another Solaris machine, and so DHCP servers that don't behave like the latter can result in unexpected behavior. The above seems to be a meaningless statement. DHCP (and its expectations) are not very complicated. I use the ISC DHCP server here, but any reasonable DHCP server should do. The DHCP client is expected to contact the server every so often and if the server fails to respond for a very long time, the DHCP lease will expire and the DHCP client should remove the IP address from the interface. However, it appears likely that the DHCP server and client are not the guilty parties here. It appears that some other software, or something in the kernel, is removing the IP address. The DHCP client expects the kernel to retain the IP address until the interface is turned down. It is not going to continually check to verify that nothing has disturbed it. Bob Just a thought: Maybe that doc (which I didn't chase down since it fell off of thread) is referencing the fact that there are special tags used for jumpstart server, image server, image name, etc. That's all I could think off: if you are trying to jumpstart a machine. And yeah, those tags do have to exist, but you can put them on an isc or any other dhcp server too. It's all just a matter of what the client is asking for and what the server replies with, the specifics of the OS don't matter, frankly. From a pure IP request/renew/accept point of view, the protocol is the same. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS on Openindiana vs ZFS on oracle linux
On 4/21/2017 2:06 PM, C. R. Oldham wrote: > For Ubuntu, it is an effort to get ZFS on the root partition. See > these wiki entries: > > https://github.com/zfsonlinux/zfs/wiki/Ubuntu > > I have several machines installed with these instructions on 16.04, > 16.10, and 17.04. They do work, but it's a pain to get it installed. > > Note that ZoL does have a couple of limitations that can make it > challenging to use it as the backing store for KVM. We have an > OpenNebula private cloud and some of qemu's caching options will cause > the VMs not to start, claiming the FS underneath doesn't expose the > right primitives. I left the OS on ZFS on those machines and > converted the partitions where the VMs keep their disks to XFS. Not to Hijack, but: I struggled with KVM on XFS for months before giving up. XFS is somewhat fundamentally incompatible with qcow2. If you are using raw, I'm sure it's great. Sine we use snapshots for backups and replication, the qcow2 functionality is a must-have. Even on a 96GB machine on a large (~40TB) qcow2 image, the fragmentation in XFS got to be enough that it was struggling to get 4MB allocations for XFS, and XFS would just fill the logs. Even echo 3> /proc/sys/vm/drop_caches and tuning linux to have at least 8MB available and use hugetlb and tuning scanning were not enough. XFS would get irrevlocably stuck and require a hard reset of the entire machine to free. You *can* tune xfs allocation sizes, but only on an empty file. You can't do it after the fact, which makes discovering that very annoying. You can also completely preallocate a qcow2, but I'm not sure it's helpful in the long run. Running KVM on ZoL on the other hand, has been wonderful so far with qcow2. (CentOS7.2) Anyway, just saying, XFS is great until it isn't, and it's failure cases will bite you months down the road and you may not have options. I'm looking forward, soon, to having another zfs machine to replicate too, then I'll use KVM+QCOW2 to take the snapshot, use zfs to replicate, then blockcommit and done. Right now it's rsync and it takes way too long. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] NSA Solaris hacks
It has been a topic of discussion on the illumos developers list today. On 4/11/2017 3:25 PM, jason matthews wrote: > > https://www.theregister.co.uk/2017/04/11/solaris_shadow_brokers_nsa_exploits/ > > > has anyone reviewed this for relevancy? > > > j. > > > ___ > 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] mbuffer connection refused on varius ports ... what to try
On 3/28/2017 5:24 AM, Harry Putnam wrote: > Timothy Coalsonwrites: > >> On Mon, Mar 27, 2017 at 9:44 PM, Harry Putnam wrote: >> >>> Geoff Nordli writes: >>> >>> [...] >>> Just a thought here, you may want to try a different ssh cipher. Give arcfour a try and see if that is fast enough for you. >>> Can that be done on ssh command line or must be done in ssh_config? >>> >>> You are transferring via mbuffer through a separate TCP connection, so >> changing the ssh cipher won't change the speed. > I think what Timothy C has in mind is dropping mbuffer/tamp and using > ssh for the whole thing > > zfs send p/fs@now | ssh RECV_HOST zfs recv p/fs > In my experience, while using mbuffer has some extra pitfalls in terms of sockets and keeping up with things when there are issues (sometimes leaving mbuffer's lying around, which was subsequently fixed in the mbuffer code with SIGCHLD), it's a lot more efficient on the network and faster to do the backups by 30%. If that isn't so important, ssh works pretty well. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] send/receive .. I expected .zfs directories
silly question: is the filesystem mounted on the receive side? if you just sent it, you'll want to mount it. Sent from my android device. -Original Message- From: Harry PutnamTo: openindiana-discuss@openindiana.org Sent: Sat, 25 Mar 2017 19:35 Subject: [OpenIndiana-discuss] send/receive .. I expected .zfs directories I just completed my very first send/receive. I was suprised to find no .zfs directories once the recv completed. Did I do something wrong or is that normal? Am I supposed to create them? Here is what I did in general. On send-host I had two fs p0/vb p0/vb/vm Sent/recv'ed p0/vb and then ditto for p0/vb/vm on recv host I tried first to just create p0. But that didn't seem to work when attempting to send p0/vb. Eventually I tried creating p0/vb on recv host... then an error message told me to use -F on send host. So: `zfs send -v p0/vb | ssh recv-host zfs recv -vF p0/vb' That worked. Then without creating anything more on recv host: `zfs send -v p0/vb/vm | ssh recv-host zfs recv -v p0/vb/vm' And that also worked. I really don't now what or how experienced people do these things. I do know that in general the docs on zfs promote using fairly fine grained fs. I assume that means as in the case above. Instead of one fs p0/vb/vm I have two as shown above... Ok when it comes to sending and receiving those fs, it would seem then that one must send first p0/vb and then p0/vb/vm If one had fs several layers deeper or with more endpoints, it seems this could get quite tedious. I suspect I may be missing the boat here somewhere but lets say it was even more fine grained. Like an fs for each vm and one for the ISOs required. p0/vb/ p0/vb/vm p0/vb/vm/vm1 p0/vb/vm/vm2 p0/vb/vm/vm3 p0/vb/vm/vm4 p0/vb/isos Then it seem this would require a whole lot of sending receiving, to move this data. Seven separate send receives by my count. So I'm convinced that either fine grainedness is not so wonderful or I'm really not understanding how this is done. I wonder how folks with some experience in this stuff, do this sort of thing? ___ 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] arp response tuning for IP Source Guard
It seems to me that you might be hitting up against "arp_defend_rate" which by default says that the maximum arps it should be expecting in one hour is 100. It's he's sending 3 per minute, that's already 180. I could be wrong. I'd probably try setting that to 300 and confirm what's going on by using "snoop arp" and then focussing in on the mac address of the switch and seeing how many are coming in an hour. On 1/6/2017 2:30 AM, the outsider wrote: Are you or the technicians SURE that this technique is still valid for current world with a lot of virtual servers attached to one switch port? This technique was invented in a time that every switchport was connected to at most one MAC address, so when the switch detects more than one MAC address at a port there was something illegal happening. (or the switchport was connected to a switch) -Oorspronkelijk bericht- Van: Tim Mooney [mailto:tim.moo...@ndsu.edu] Verzonden: vrijdag 6 januari 2017 0:50 Aan: openindiana-discuss@openindiana.org Onderwerp: Re: [OpenIndiana-discuss] arp response tuning for IP Source Guard In regard to: Re: [OpenIndiana-discuss] arp response tuning for IP Source...: On 01/05/17 15:37, Tim Mooney wrote: When that was enabled for the subnet I'm on, my hipster workstation and the hipster VirtualBox VM I have both started experiencing packet loss. Talking with the network engineers, the Cisco switch is sending batches of 3 ARP probes periodically, and both my workstation and the VM appear to be periodically not responding to the ARP probes. That causes the switch to temporarily ban/block packets from either system, which is what's causing the intermittent packet loss. Anyone have any suggestions for what tuning I should be looking at that would tell the Illumos network stack that it's OK to respond to semi-frequent batches of ARP probes? It would be great to see the syslog messages and (if possible) a packet trace showing what's going on. In general, if the system itself is directly responsible for these outages, it will at least log something about the event. At the log level I've been running at, there hasn't been anything useful logged related to this. If necessary, I can definitely dial up the logging. Are these ARP requests or responses? There are subtle differences between the two. According to our principal network engineer, the Cisco switch was defaulting to sending 3 ARP probes (in quick succession) every 60 seconds. He has since dialed that back to just 1 per 60 seconds for this particular switch, to see if that had any impact on the issue, but it did not. He's done a bunch more research since I sent my initial question to this list, and right now he thinks the issue may be that the ARP probe from the Cisco switch is unicast, but Solaris apparently may be issuing ARP responses as *broadcast*, which the switch may not be expecting. The reference he found related to broadcast ARP responses is here: http://seclists.org/nmap-dev/2009/q1/176 http://unix.derkeiler.com/Mailing-Lists/SunManagers/2009-01/msg00015.html He's also suggested that I might be able to set 'arp_defend_interval' to something like 20 seconds, so that my workstation just periodically sends unsolicited ARPs for itself, to essentially preempt the switch's probes. Based on the docs he found: http://docs.oracle.com/cd/E36784_01/html/E36845/gnogz.html Since the docs say "Never" in answer to the "When to change" for any of these settings, I haven't actually tried setting arp_defend_interval. The way I read the docs, it seems like arp_publish_interval might be better, but I know better than to argue with our principal network engineer about anything network related. :-) Based on what I remember from working on this code many years ago, one of the really confusing bits to deal with is Ethernet bridge ("switch") behavior itself. Many bridges (I think at least Extreme, and probably others) have special mechanisms built-in to protect against ARP storms, and they rate-limit based on the number of broadcasts. This is (I believe!) independent of any sort of "Source Guard" feature. I ran into this issue numerous times when testing Solaris IP Duplicate Address Detection. Thanks much for the response! Tim ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org https://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] using touch ... whats wrong here
That is the classical answer for ctime. James is correct. However some may find it interesting to note that GPFS does indeed keep the original creation time as an additional attribute that can then be used for policy applications. It is not exposed to Unix, but it's there. Alas, GPFS is only available on AIX and Linux at this time. (misc morning info share) Sent from my android device. -Original Message- From: James CarlsonTo: openindiana-discuss@openindiana.org Sent: Thu, 03 Dec 2015 8:24 Subject: Re: [OpenIndiana-discuss] using touch ... whats wrong here On 12/03/15 04:56, Peter Tribble wrote: > You're mixing the various timestamps. Indeed; that's the problem. > access, shown by ls -lu > modify (of data), shown by ls -l > creation (or modify metadata), shown by ls -lc Perhaps a nit, but one I think is important. UNIX (including OpenIndiana) does not keep track of file or directory "creation" time. The 'ctime' element tracks the last attribute change time. There's no way (in UNIX) of knowing when a file was created. And, yes, the OP should be using the file's mtime (the "-newer" option). The attribute change time is rarely what you want. :-/ -- James Carlson 42.703N 71.076W ___ 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] What is the recommended way to back up root pool?
I'd put that in my config mgmt system (or installer) Sent from my android device. -Original Message- From: Bob Friesenhahn <bfrie...@simple.dallas.tx.us> To: Discussion list for OpenIndiana <openindiana-discuss@openindiana.org> Sent: Wed, 28 Oct 2015 21:00 Subject: Re: [OpenIndiana-discuss] What is the recommended way to back up root pool? On Wed, 28 Oct 2015, Doug Hughes wrote: > for home or for office? > for office, I don't back up root pool. it's considered disposible and > reproducible via reinstall. (that plus config management) > for home, you can zfs send it somewhere to a file if you want, or you can > tar it up since that's probably easier to restore individual files after an > oops. I do the latter. Then you could reinstall from golden image and > restore the files you need. Assume that this is for a network server with advanced network configuration settings, ssh config, zones, etc. If was the same as a standard OS install without subsequent configuration, then backing up would not be so important. Bob -- Bob Friesenhahn bfrie...@simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer,http://www.GraphicsMagick.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] What is the recommended way to back up root pool?
for home or for office? for office, I don't back up root pool. it's considered disposible and reproducible via reinstall. (that plus config management) for home, you can zfs send it somewhere to a file if you want, or you can tar it up since that's probably easier to restore individual files after an oops. I do the latter. Then you could reinstall from golden image and restore the files you need. On Wed, Oct 28, 2015 at 7:24 PM, Bob Friesenhahn < bfrie...@simple.dallas.tx.us> wrote: > What is the recommended approach to back up a zfs root pool? > > For other pools I use zfs send/receive and/or rsync-based methods. > > The zfs root pool is different since it contains multiple filesystems, > with the filesystem for one one BE being mounted at a time: > > % zfs list -r -t filesystem rpool > NAMEUSED AVAIL REFER MOUNTPOINT > rpool 79.7G 377G50K /rpool > rpool/ROOT 29.5G 377G31K legacy > rpool/ROOT/openindiana 15.8M 377G 3.15G / > rpool/ROOT/openindiana-1 38.9M 377G 5.97G / > rpool/ROOT/openindiana-2 40.6M 377G 11.3G / > rpool/ROOT/openindiana-2-backup-1 124K 377G 10.5G / > rpool/ROOT/openindiana-3 48.4M 377G 13.2G / > rpool/ROOT/openindiana-3-backup-176K 377G 11.4G / > rpool/ROOT/openindiana-3-backup-245K 377G 11.5G / > rpool/ROOT/openindiana-3-backup-3 123K 377G 11.9G / > rpool/ROOT/openindiana-3-backup-444K 377G 12.1G / > rpool/ROOT/openindiana-4 20.1M 377G 18.8G / > rpool/ROOT/openindiana-4-backup-195K 377G 13.3G / > rpool/ROOT/openindiana-4-backup-2 156K 377G 18.8G / > rpool/ROOT/openindiana-5 17.5M 377G 18.9G / > rpool/ROOT/openindiana-6 29.4G 377G 17.6G / > rpool/export121M 377G32K /export > rpool/export/home 121M 377G32K /export/home > > This means that there are multiple filesystems which would need to be > backed up in order to save a replica of the pool. > > At the moment I am using rsync-based backup of only selected filesystems. > > It is not clear to me where configuration due to utilities like 'ipadm' > and 'dladm' is stored, but I am pretty sure it is to files under the /etc > directory. > > What is recommended/common practice for backing up the root pool? > > Bob > -- > Bob Friesenhahn > bfrie...@simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ > GraphicsMagick Maintainer,http://www.GraphicsMagick.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] Antwort: Transferring everything in one zpool to be a, subtree in another larger zpool (Hans J. Albertsson)
you can also use ttcp or iperf just fine. I do that all the time and they are pretty much available anywhere. (ttcp is so simple it's just a single .c source file). The buffering isn't really the important part, using as close to raw tcp transport gets you the biggest benefit (vs encryption/decryption overhead and slowdown). Sent from my android device. -Original Message- From: Hans J. Albertsson hans.j.alberts...@gmail.com To: Discussion list for OpenIndiana openindiana-discuss@openindiana.org Sent: Tue, 05 May 2015 9:31 Subject: Re: [OpenIndiana-discuss] Antwort: Transferring everything in one zpool to be a, subtree in another larger zpool (Hans J. Albertsson) mbuffer isn't included in FreeNAS. It IS available for FreeBSD, but I don't know if that version will behave responsibly in FreeNAS. FreeNAS devs say mbuffer won't be included in FreeNAS because we already do proper buffering via dd. Would dd work for buffering in Solaris/OI?? On 2015-05-05 14:26, Sebastian Gabler wrote: Am 05.05.2015 um 14:00 schrieb openindiana-discuss-requ...@openindiana.org: As a side note; just piping to ssh is excruciatingly slow, using netcat, nc, speeds things up at least 4fold. The method that worked fine for me was using mbuffer. (I couldn't make netcat work, shame on me, but I also remember that there were several pros of using mbuffer. I think one of them was the ability to use digest to secure the transfer.) See: http://everycity.co.uk/alasdair/2010/07/using-mbuffer-to-speed-up-slow-zfs-send-zfs-receive/ That being said, it's exactly two years when I last did a migration with zfs send | zfs receive. I hope that still works. BR, Sebastian ___ 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] data on rpool - Bad idea?
The reason that data volumes on rpool is generally not a good idea is for recoverability. You can take all of the disks of a given pool and move them to another system, except rpool. rpool defines the system itself, so data volumes there are tied to that system. Data volumes in any other pool can move between systems very easily, which is great when you have certain classes of hardware failures, or you want to upgrade your system, or something similar. The data is thus independent of the current disposition of the OS. At some point later on, you'll likely regret having done it. Sent from my android device. -Original Message- From: Harry Putnam rea...@newsguy.com To: openindiana-discuss@openindiana.org Sent: Sat, 07 Feb 2015 7:24 Subject: [OpenIndiana-discuss] data on rpool - Bad idea? I've read that it is a bad idea to have datasets on the same disc that rpool is on. I'm not sure why that is so. In my current situation I need more space but to rig up my `HP xw8600 workstation' with more disc space will require getting a cage and some new discs, not to mention the huffing and puffing necessary to get it all installed and usable. When I installed oi on this hardware I thought it was a smart idea to use 2 older 250 GB discs for a 2 disc mirrored setup of rpool. And a pair of 1tb discs for 2disc mirror of datasets. At that time I hadn't really realized what it would take to add discs. But of course I'm now running out of space... and still not flush enough to pop for the $400 or so to add a cage and 4 2-3 tb discs. However I do have 2 1tb discs laying unused. So, (finally cutting to chase) how bad of an idea would it be to use those two 1tb discs for a mirrored rpool and whatever data. I would gain about 800-850 GB of mirrored data space but it would involve putting data on same discs and rpool. (I'll leave for another thread to discuss how to get rpool onto the new discs in the most painless manner) So, what are your best thoughts on employing the two 1tb discs as described above? ___ 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] A ZFS related question: How successful is ZFS, really???
Couple of points and counter points from my own experience. *) tape really isn't dead. No, really. at about $.01/GB/copy, and 1x10^20 bit error rate, you can't beat it. Use it for the right thing though. This excels as an offline archival media with media lifetimes expected at around 30 years. Contrast that to cheap disk. Everybody forgets that you still have to account for the chassis, memory, space, transportability, media bandwidth, brackets, cpus, and power used by an online storage system. Also, if you go with the really inexpensive disks to keep costs down, you sacrifice 5 orders of magnitude of media reliability when compared to tape. Seriously. There's a place for both. Places serious about data integrity and offsite archive at good cost still use tape. it's not dead. It's not even dying contrary to what some popular reading might say in some places. You can backup zfs to tape using zfs diff to get your list and then whatever mechanism you want to get those items to tape (rsync, tsm, whatever). The point is, that full-filesystem-scan for backups is definitely dead. Don't do that. There are better mechanisms. Full scan doesn't scale. *) ZFS is really cool and all. We use it a lot! To say it is unquestionably the best might be a bit of an overstatement. GPFS is also really cool, and arguably better in many ways. The erasure coding in the new de-clustered RAID is better than ZFS in terms of rebuild time and certain rare causes of data loss. GPFS allows easy separation of metadata onto fast storage for index and search. The GPFS policy engine is way cool. You can arbitrarily relayout your GPFS across storage. You can have pools of different characteristics and do policy migration or data placement policies arbitrarily to use them. (We STILL can't do any relayout of any kind on ZFS. Argh). The big downside is price, of course. There are plenty of free ZFS solutions available but none for GPFS. BTRFS has a technological edge over zfs in relayout as well. Why isn't it possible to shrink/relayout ZFS still? IMHO this should have been delivered about 2 years ago. As far as stability and marketshare go, ZFS gets the big win, naturally. On Mon, Jan 12, 2015 at 9:56 AM, Hans J Albertsson hans.j.alberts...@gmail.com wrote: Does anyone have anything beyond own impressions and war stories? Is anyone collecting statistics on storage solutions sold? Hans J. Albertsson From my Nexus 5 Den 12 jan 2015 15:24 skrev Schweiss, Chip c...@innovates.com: On Mon, Jan 12, 2015 at 8:17 AM, Andrew Gabriel illu...@cucumber.demon.co.uk wrote: Since you mention Sun/Oracle, I don't see them pushing ZFS very much anymore, although I am aware their engineers still work on it. Oracle pushes ZFS hard and aggressively. I dare you to fill out their contact form or download their virtual appliance demo. Their sales people will be calling within the hour. We just recently went through a bidding war on an HA + DR system with 1/2 PB useable storage with many vendors including Nexenta and Oracle. Oracle was price competitive with Nexenta and is in my opinion a much more polished product. We still chose to build our own on OmniOS because we could still do that for about 1/2 the price of Oracle / Nexenta. That's less than 1/4 the price of 3par/HP, IBM, Dell/Compellent BTW, our OmniOS build is on the exact same hardware, Nexenta would have been. -Chip ___ 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] How to revert a permissive chmod
Well, it's not the exactly same, but you could do chmod -R go-w . On 1/11/2015 3:33 PM, Harry Putnam wrote: Is there some similarly all incompassing command to revert: /bin/chmod -R A=everyone@:full_set:fd:allow /some-dir back to a more default chmod 755 on directories and 644 on type -f files? (and no acls set) I mean short of using `find' in comination with chmod. ___ 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] How to revert a permissive chmod
You could use A= with -R as long as you know the number of the item that you are trying to modify (visible with -v or -V depending on your viewing pleasure) It's probably the 'everyone' one... On 1/11/2015 4:21 PM, Harry Putnam wrote: Doug Hughes d...@will.to writes: Well, it's not the exactly same, but you could do chmod -R go-w . Has no effect on all the thousands of type -f files ___ 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] destroy all but specific snapshot
On 9/30/2014 8:31 PM, Tim Aslat wrote: Harry Putnam wrote on 01/10/2014 09:52: This is not so easy to find in google searches How does one go about destroying all but a specific snapshot? The one I want is somewhere in the middle timewise So not wanting to use `destroy -r'. This is in a .zfs/snapshot/* where there are many auto snaps. So looking for a way besides destroying -r or 1 by 1. Hi Harry, Simplest option would be something like zfs list -t snapshot -o name -H -r filesystem/path /tmp/snaplist.txt edit /tmp/snaplist.txt to remove the snapshot you want to keep for SNAP in `cat /tmp/snaplist.txt` do zfs destroy $SNAP done Here's an equivalent one liner for you. zfs list -t snapshot -H pool | grep -v the one you want to exclude | awk '{print $1}' | xargs -n 1 zfs destroy (you should run up to the last pipe and confirm that's what you want before adding on the xargs. ;) * warning: if you have replication to a backup server, removing all of the snapshots for other filesystems will make those replicated snapshots non-operative. ___ openindiana-discuss mailing list openindiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] fuiture small expansion of raidz1 possible?
On 2/11/2014 5:57 PM, Brogyányi József wrote: And on a similar note: Suppose I have a signgle disk with data, and I decide I can just afford a raidz1 of 4 disks, so I buy 3 more can I somehow migrate the data on the original single disk onto a 3 disk raidz1 and then add the original disk to that raidz? You do realize that this will be effectively a RAID-0 reliability with RAID-5 performance? It's an interesting concept, but I'd be very worried about a single block error causing data loss and real problems. Maybe your window is short enough that you don't mind the risk. Hi Here is the technik: http://www.dev-eth0.de/creating-raidz-with-missing-device/ It must work. Brogyi ___ 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] Powerloss protected SSDs for ...Re: Low low end server
Why not Intel 320 series? Also 710 series work fine for this, for a bit more $$ and a bit more speed. The 320 are not as fast as the S3700 or S3500 but they are a LOT less expensive. On Mon, Feb 10, 2014 at 4:32 PM, Schweiss, Chip c...@innovates.com wrote: On Mon, Feb 10, 2014 at 5:22 AM, Hans J. Albertsson hans.j.alberts...@branneriet.se wrote: Samsung 843 The 843 while called and enterprise SSD, does not have capacitors for power loss protection. http://www.thessdreview.com/our-reviews/samsung-843/2/ ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Powerloss protected SSDs for ...Re: Low low end server
true, Volker.. Just to note though, the 320s have no battery, but they do have enough capacitor to flush anything from the small ram into flash on power outage. On Mon, Feb 10, 2014 at 4:58 PM, Volker A. Brandt v...@bb-c.de wrote: Why not Intel 320 series? Also 710 series work fine for this, for a bit more $$ and a bit more speed. The 320 are not as fast as the S3700 or S3500 but they are a LOT less expensive. This thread started out as a discussion of the merits of the HP N54L microserver for home use. I am not really sure if a home server needs mirrored battery-protected SSDs. :-) Regards -- Volker -- Volker A. Brandt Consulting and Support for Oracle Solaris Brandt Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANYEmail: v...@bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgröße: 46 Geschäftsführer: Rainer J.H. Brandt und Volker A. Brandt When logic and proportion have fallen sloppy dead ___ 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] Low low end server
smartctl -x reports the wear on decent ones (read: you shouldn't consider any that doesn't have this feature). When it gets close to 0, or you see a lot of errors, it's time to replace it. On Mon, Feb 10, 2014 at 5:00 PM, Brogyányi József bro...@gmail.com wrote: 2014.02.09. 22:19 keltezéssel, Jim Klimov írta: 2*120Gb Samsung Pro SSDs (with powerloss protection, formatted to use 100Gb for mirrored rpool/mirrored zil/striped l2arc since their 100Gb sub-model has much higher reliability and speed - but is not on sale here). Jim Only one question about the SSD usage. How do I know the SSD status? You know the SSD capacity is decreasing. When have to replace it? Regards Brogyi ___ 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] Powerloss protected SSDs for ...Re: Low low end server
capacitors are better. Batteries wear out and are difficult to have the correct monitoring for replacement. On Mon, Feb 10, 2014 at 5:51 PM, Ben Taylor bentaylor.sol...@gmail.comwrote: Surprised someone hasn't developed a SATA power cable with small battery Passthrough for this exact application. On Mon, Feb 10, 2014 at 5:06 PM, Doug Hughes d...@will.to wrote: true, Volker.. Just to note though, the 320s have no battery, but they do have enough capacitor to flush anything from the small ram into flash on power outage. On Mon, Feb 10, 2014 at 4:58 PM, Volker A. Brandt v...@bb-c.de wrote: Why not Intel 320 series? Also 710 series work fine for this, for a bit more $$ and a bit more speed. The 320 are not as fast as the S3700 or S3500 but they are a LOT less expensive. This thread started out as a discussion of the merits of the HP N54L microserver for home use. I am not really sure if a home server needs mirrored battery-protected SSDs. :-) Regards -- Volker -- Volker A. Brandt Consulting and Support for Oracle Solaris Brandt Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANYEmail: v...@bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgröße: 46 Geschäftsführer: Rainer J.H. Brandt und Volker A. Brandt When logic and proportion have fallen sloppy dead ___ 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] real-time syncing two directories across servers (unison ?)
On 12/11/2013 8:13 PM, Edward Ned Harvey (openindiana) wrote: From: Gregory Youngblood [mailto:greg...@youngblood.me] Check out owncloud. The open source components might be useful. I personally, and two other IT guys that I've spoken with from different companies, have been burned by placing any trust in owncloud. In fact, I'm still subscribed to their mailing list, and at least once a day, new people write in, asking for help, because it either fails to sync files it should sync, or some data seems to be lost. My advice: Steer clear. In Microsoft, what you're looking for is called DFS. I don't know any good linux/unix based bidirectional tools such as requested, but if you start searching for alternatives to DFS (or if you simply pay MS and deploy DFS) then you have an answer, and hopefully can find some FOSS solution out there. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss Seafile seems to work reasonably well. We've actually been looking for something with more capability though, one that can handle various levels of permissions and multiple-projects. But, if all you want is something to plugin to your own website with Dropbox like capabilities and multi-device synchronization capability, Seafile seems to be ready to fill that bit. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] VMware
We have lacp working between force10, hp, and cisco switches in all possible combinations with no difficulties. We do monitor and alert on excessive errors and drops for interfaces, but lacp isnt a culprit. If anything, it's an underlying interface when we find them. Also, it beats the heck out of spanning tree and is 2 orders of magnitude simpler than ospf, and 1 order simpler and more portable than ecmp. I am quite surprised by your observations. Sent from my android device. -Original Message- From: Edward Ned Harvey (openindiana) openindi...@nedharvey.com To: Discussion list for OpenIndiana openindiana-discuss@openindiana.org Sent: Tue, 13 Aug 2013 7:22 AM Subject: Re: [OpenIndiana-discuss] VMware From: James Relph [mailto:ja...@themacplace.co.uk] Sent: Monday, August 12, 2013 4:47 PM No, we're not getting any ping loss, that's the thing. The network looks entirely faultless. We've run pings for 24 hours with no ping loss. Yeah, I swore you said you had ping loss before - but if not - I don't think ping alone is sufficient. You have to find the error counters on the LACP interfaces. Everybody everywhere seems to blindly assume LACP works reliably, but to me, simply saying the term LACP is a red flag. It's extremely temperamental, and the resultant behavior is exactly as you've described. ___ 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] Sun Fire
Most of the early marvell problems have been addressed with software fixes in the oracle release of solaris. I couldnt say about OI. Yes, the 4540 is better all around. Sent from my android device. -Original Message- From: Ben Taylor bentaylor.sol...@gmail.com To: Discussion list for OpenIndiana openindiana-discuss@openindiana.org Sent: Fri, 21 Jun 2013 1:34 PM Subject: Re: [OpenIndiana-discuss] Sun Fire x4540s use the LSI controller. We got an upgrade on our x4500 because of systemic problems with our Raid-z setup and the Marvel controllers and the drivers (This has been several years, maybe 4) On Fri, Jun 21, 2013 at 4:48 PM, Gary Gendel g...@genashor.com wrote: On 06/21/2013 11:56 AM, Peter Tribble wrote: On Fri, Jun 21, 2013 at 1:58 PM, Gary Gendel g...@genashor.com wrote: I have a possibility of picking up a decommissioned X4340 (thumper) Hm. An X4340 is some sort of power transfer unit. Thumper = X4500 Thor = X4540 Whoops, it's the X4540 w/o disks That's the killer. These boxes are very fussy about the drives that work; the real value of them now is for the disks they contain, you're going to have trouble getting drives (and it's going to cost). Interesting. I thought they used the same Marvell MV88SX6081 8-port controllers I picked up for the v20z. I've been using that for modern WD and Seagate SATA drives without issue. to replace my aging V20z ( and multiple external drive stacks cheap. It looks like it should be supported by looking at the HCL. Is there any gotchas I should be aware of before I commit to purchase? Is there something more modern I should be looking at instead? Probably look at something more modern, and something sized to suit. It really needs to be full of 48 drives, only a couple of models are supported, and they're all getting a bit old. Just looking for something cheap and big enough for at least 8 drives. I currently have two 5-drive external cases. Since I haven't gotten the sata multiplexers to work, I have all 8 sata cables running out of the marvell controller directly to each drive, my own personal octopus. Can't get much more of a kludge than that. Gary __**_ OpenIndiana-discuss mailing list OpenIndiana-discuss@**openindiana.orgOpenIndiana-discuss@openindiana.org http://openindiana.org/**mailman/listinfo/openindiana-**discusshttp://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] Recommendations for fast storage (OpenIndiana-discuss Digest, Vol 33, Issue 20)
some of these points are a bit dated. Allow me to make some updates. I'm sure that you are aware that most 10gig switches these days are cut through and not store and forward. That's Arista, HP, Dell Force10, Mellanox, and IBM/Blade. Cisco has a mix of things, but they aren't really in the low latency space. The 10g and 40g port to port forwarding is in nanoseconds. buffering is mostly reserved to carrier operations anymore, and even there it is becoming less common because of the toll it causes to things like IPVideo and VOIP. Buffers are good for web farms, still, and to a certain extent storage servers or WAN links where there is a high degree of contention from disparate traffic. At a physical level, the signalling of IB compared to Ethernet (10g+) is very similar, which is why Mellanox can make a single chip that does 10gbit 40gbit, and QDR and FDR infiniband on any port. there are also a fair number of vendors that support RDMA in ethernet NIC now, like SolarFlare with Onboot technology. The main reason for lowest achievable latency is higher speed. Latency is roughly equivalent to the inversion of bandwidth. But, the higher levels of protocols that you stack on top contribute much more than the hardware theoretical minimums or maximums. TCP/IP is a killer in terms of adding overhead. That's why there are protocols like ISER, SRP, and friends. RDMA is much faster than the kernel overhead induced by TCP session setups and other host side user/kernel boundaries and buffering. PCI latency is also higher than the port to port latency on a good 10g switch, nevermind 40 or FDR infiniband. There is even a special layer that you can write custom protocols to on Infiniband called Verbs for lowering latency further. Infiniband is inherently a layer1 and 2 protocol, and the subnet manager (software) is resposible for setting up all virtual circuits (routes between hosts on the fabric) and rerouting when a path goes bad. Also, the link aggregation, as you mention, is rock solid and amazingly good. Auto rerouting is fabulous and super fast. But, you don't get layer3. TCP over IB works out of the box, but adds large overhead. Still, it does make it possible that you can have IB native and IP over IB with gateways to a TCP network with a single cable. That's pretty cool. Sent from my android device. -Original Message- From: Edward Ned Harvey (openindiana) openindi...@nedharvey.com To: Discussion list for OpenIndiana openindiana-discuss@openindiana.org Sent: Tue, 16 Apr 2013 10:49 AM Subject: Re: [OpenIndiana-discuss] Recommendations for fast storage (OpenIndiana-discuss Digest, Vol 33, Issue 20) From: Bob Friesenhahn [mailto:bfrie...@simple.dallas.tx.us] It would be difficult to believe that 10Gbit Ethernet offers better bandwidth than 56Gbit Infiniband (the current offering). The swiching model is quite similar. The main reason why IB offers better latency is a better HBA hardware interface and a specialized stack. 5X is 5X. Put another way, the reason infiniband is so much higher throughput and lower latency than ethernet is because the switching (at the physical layer) is completely different from ethernet, and messages are passed directly from user-level to user-level on remote system ram via RDMA, bypassing the OSI layer model and other kernel overhead. I read a paper from vmware, where they implemented RDMA over ethernet and doubled the speed of vmotion (but still not as fast as infiniband, by like 4x.) Beside the bypassing of OSI layers and kernel latency, IB latency is lower because Ethernet switches use store-and-forward buffering managed by the backplane in the switch, in which a sender sends a packet to a buffer on the switch, which then pushes it through the backplane, and finally to another buffer on the destination. IB uses cross-bar, or cut-through switching, in which the sending host channel adapter signals the destination address to the switch, then waits for the channel to be opened. Once the channel is opened, it stays open, and the switch in between is nothing but signal amplification (as well as additional virtual lanes for congestion management, and other functions). The sender writes directly to RAM on the destination via RDMA, no buffering in between. Bypassing the OSI layer model. Hence much lower latency. IB also has native link aggregation into data-striped lanes, hence the 1x, 2x, 4x, 16x designations, and the 40Gbit specifications. Something which is quasi-possible in ethernet via LACP, but not as good and not the same. IB guarantees packets delivered in the right order, with native congestion control as compared to ethernet which may drop packets and TCP must detect and retransmit... Ethernet includes a lot of support for IP addressing, and variable link speeds (some 10Gbit, 10/100, 1G etc) and all of this asynchronous. For these reasons, IB is not a suitable replacement for IP
Re: [OpenIndiana-discuss] Tape backup
On 3/17/2013 6:23 PM, Reginald Beardsley wrote: Tape as an archival medium has significant issues. Reading poorly stored tapes is a one try proposition w/ no assurance of success. The first high volume commercial application for digital tape was seismic data acquisition for the oil industry. The oil companies had very detailed cleaning and retensioning schedules w/ a large staff to perform them on the tape archives. Absent that level of care, reading old tapes is very difficult and requires great skill. Old tape is NOT fun to work with. High capacity tape drives and tapes are not cheap either. Blank LTO tape is almost as expensive as SATA disk. A ZFS based remote replicating server using triple parity RAIDZ is probably cheaper than tape. For extremely large volumes and long archival periods, optical tape is probably the best choice. But then you're probably working for the government. I would strongly urge comparing the cost of a ZFS backup server w/ daily snapshots to the cost of conventional tape backup. I think you'll be quite surprised at the implications. reading old disks is a just as significant if not more of an issue, by my estimation. Try to find a machine that you can do low voltage differential disks with these days. That was only 15 years ago. what about SMD? The controllers keep changing over time. Also, after a disk has been in use for a significant period of time (say years), the lubrication on the platters tends to evaporate a little bit so that when you leave it off for a long period of time (days/weeks) it will stick to the heads and the platters. Tape doesn't have that issue. An LTO5 tape is about $30 each, better in quantity. 1.5TB, more depending upon compression. That's an enterprise quality tape with much longer shelf life than a cheap deskstar disk if cared for properly. Even the cheap 1TB disks are $70. That's almost a 3X advantage. There's still a place for tape for archival, and yes, you do have to care for it properly, just like you have to care for everything else. But, cared for properly, tape should still outlive disk. But, used it in its proper place. zfs snapshots make a lot of sense for online backups! ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] multiple IP addresses, same NIC
On 3/5/2013 10:18 PM, Edward Ned Harvey (openindiana) wrote: From: Doug Hughes [mailto:d...@will.to] 2) explicitly set the route for 192.168.10.x : route add 192.168.10.0/mask 192.168.2.1 That's what I'm saying I have already done. I set the default route to 192.168.1.1, and I set a static route, 192.168.10.x/24 via 192.168.2.1. The route is in effect, as evidenced: For simplicity, let's say 192.168.1.1 has MAC 11:11:11:11:11:11 and let's say 192.168.2.1 has mac 22:22:22:22:22:22. When I ping something on the internet, I see a packet go out my NIC, source IP 192.168.1.100, destination MAC 11:11:11:11:11:11 and destination IP 8.8.8.8. It all works, I get a ping response. When I ping 192.168.2.1 directly, I see a packet go out my NIC, source IP 192.168.2.100, destination MAC 22:22:22:22:22:22 and destination IP 192.168.2.1. It all works, I get a ping response. When I ping something on the other end of the VPN, I see a packet go out of my NIC, source IP 192.168.1.100, destination MAC 22:22:22:22:22:22 and destination IP 192.168.10.10 (or whatever.) The firewall drops the packet, because duh, the source IP isn't in the same subnet as the firewall. I am also exploring the NAT option, assuming I'm not going to be able to resolve the above problem. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss is it the same if you use something other than ping? Does it work correctly if you use ping -i to specify the source interface? Another option may be tagged vlans. Supporting a separate tagged interface would be very clear for the host. It would appear as a totally separate layer2 so there'd be no chance for mac reuse. That's only if your router can support that too, however... ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] OI Crash
On 1/18/2013 7:53 PM, dormitionsk...@hotmail.com wrote: On Jan 17, 2013, at 8:47 PM, Reginald Beardsley wrote: As far as I'm concerned, problems like this are a bottomless abyss. Which is why I'm still putting up w/ my OI box hanging. It's annoying, but not critical. It's also why critical stuff still runs on Solaris 10. Intermittent failures are the worst time sink there is. There is no assurance that devoting all your time to the problem will fix it even at very high skill levels w/ a full complement of the very best tools. If you're getting crash dumps there is hope of finding the cause, so that's a big improvement. Good luck, Reg BTW Back in the 80's there was a VAX operator in Texas who went out to his truck, got a .357 and shot the computer. His employer was not happy. But I can certainly understand how the operator felt. From 1992 to I used to 1998, I used to work at the Denver Museum of Natural History -- now the Denver Museum of Nature and Science. We had two or three DEC Vax's and an AIX machine there. It was their policy that once a week we had to power each of the servers all the way down to clear out any memory problems -- or whatever -- as preventive maintenance. Since then, I've always had the habit of setting up a cron job to reboot my servers once a week. It's not as good as a full power down, but it's better than nothing. And in all these years, I've never had to deal with intermittent problems like this, except for a few brief times when I used Red Hat Linux ten plus years ago. (I've tried most of Red Hat's versions since 6.2, and RHEL 6 is the first version I've found that runs decent enough on our hardware, and that I'm happy enough with, for us to use.) So, if you can do it, you might want try setting up a cron job to reboot your server once a week -- or every night. I reboot our LTSP thin client server every night just because it gets hit with running lots of desktop applications that I think give it a greater potential for these kinds of memory problems. On the other hand, we have all of our websites hosted on one of our parishioner's servers -- and he doesn't reboot his machines periodically like I do -- and about every two months, I have to call him up and tell him something is wrong. And he goes and powers down his system -- sometimes he has to even unplug it -- and then turn it back on, and everything works again. I know there are system admins that just love to brag about how great their up-times are on their machines -- but this might just save you a lot of time and grief. Of course, if you're running a real high-volume server, this might not be workable for you; but it only takes 2-5 minutes or so to reboot... Perhaps in the middle of the night you might be able to spare it being down that short time? Just a friendly suggestion. Shared experience. I know others may tell you that that's no longer necessary anymore in these more modern times; but my experience has been otherwise. I hope it helps. +Peter, hieromonk Haven't we passed the days of mystical sysadmin without understanding and characterization? Keeping up tradition for tradition's sake without understanding the underlying reasons really doesn't do anybody a favor. If there are memory leaks, we posses the technology to find them. My organization has thousands of machines that run jobs sometimes for months at a time. If I had to reboot servers once a week, my users would be at the doors with pitchforks. The only time we take downtime is when there are reasons to do so, including OS updates, hardware failures, and user software run amok. They can run a very long time like this. Not that memory leaks never happen. Of course they do, but they eventually get found and fixed, or the program causing them passes into obsolescence. Always. I encourage discovery rather than superstition, and diagnosis rather than repetition. Be a knight, not a victim! ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Finding pools after reinstall
That isn't necessarily a driver. What model hba do you have? If it is among many versions of megaraid or similar, you won't see any disks until you create logical units from the raid card bios interface. On Dec 19, 2012 9:53 AM, Thommy M. Malmström thommy.m.malmst...@gmail.com wrote: Sorry I didn't read this thread till today - The first thing is, you *must* be able to see the disks via sudo format -e /dev/null If they're not there, reboot, make sure they're enabled in BIOS, and sudo devfsadm -Cv After that, they should be visible, or else you're missing a driver. I'm definitely missing a driver. Can't see the disks with format. Will test Solaris 11.1 now to see if it finds all my disks. Thanks for response. /Thommy M. ___ 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] Finding pools after reinstall
Have you tried to import the pools by name? Often that will work where discovery isn't so great. On 12/9/2012 1:35 PM, Thommy M. Malmström wrote: Gentlemen, you have to be really careful with me here. It's some years since I've touched Solaris. :) I have a file server with several disks and OpenSolaris b125 (iirc). It has hummed away nicely since 2009 or so and I've done nothing except for changing a faulty disk drive. The other day it went down and refused to boot. I got a new disk and disconnected all drives and replaced the boot disk. Then I installed OpenIndiana 147 on the new disk and reconnected the other disks. BUT, when I try to find the zpools on the old disk with zpool import, it says nada. And format only lists one disk besides the boot disk. So, ideas? /Thommy M. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS High-Availability and Sync Replication
On Tue, Nov 20, 2012 at 7:01 AM, Edward Ned Harvey (openindiana) openindi...@nedharvey.com wrote: From: Doug Hughes [mailto:d...@will.to] Well, to me, the most obvious is use another box with ZFS to mirror the ISCSI devices on ZFS. I'm in the process of doing this myself with some stuff. 2 head nodes and 2 storage nodes. The storage nodes are targets and the head nodes are initators. You get performance and reliability, and failover between head nodes is relatively easy with zpool import. This is precisely the thing I did with poor results, except that I didn't have 4 nodes - I had only 2. So each node was both an initiator and a target. I made the initiator service dependent on the target service (so the initiator would go down first and come up second). I made zpool import of the iscsi pool dependent on the initiator being up, and I made vbox dependent on the iscsi zpool mounter. (see simplesmf, http://code.google.com/p/simplesmf/ for the zpool dependent on iscsi initiator) My word of advice, until such time (which may never come) that I might know more... Never, ever let the iscsi targets go down unless the initiators are both down first. strange. I tested failing the targets while the initiators are up and failover worked fine. One half of the mirror went away and everything worked as expected. I wonder if you might be running into some race or other strange locking condition in the ISCSI kernel code being both target and initiator for the same mirrored zpool? I tested it pretty thoroughly. Aside from needing to run tcpkill to make Linux NFS clients really give up on the current NFS/TCP connection and reestablish (or wait 5 minutes), things worked as expected. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS High-Availability and Sync Replication
On 11/19/2012 9:39 PM, Edward Ned Harvey (openindiana) wrote: From: Sašo Kiselkov [mailto:skiselkov...@gmail.com] I've been lately looking around the net for high-availability and sync replication solutions for ZFS and came up pretty dry - seems like all the jazz is going around on Linux with corosync/pacemaker and DRBD. I found a couple of tools, such as AVS and OHAC, but these seem rather unmaintained, so it got me wondering what others use for ZFS clustering, HA and sync replication. Can somebody please point me in the right direction? Despite my recent failure, I still think there must be a good solution mirroring an iscsi device with another iscsi device. I've put this project on the back burner, but I'm still optimistic that there's a simple explainable and repeatable problem causing the failure I experienced. (And avoidable.) Well, to me, the most obvious is use another box with ZFS to mirror the ISCSI devices on ZFS. I'm in the process of doing this myself with some stuff. 2 head nodes and 2 storage nodes. The storage nodes are targets and the head nodes are initators. You get performance and reliability, and failover between head nodes is relatively easy with zpool import. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS remote receive
On 10/23/2012 7:52 AM, Sebastian Gabler wrote: Hi, I am facing a problem with zfs receive through ssh. As usually, root can't log on ssh; the log on users can't receive a zfs stream (rights problem), and pfexec is disabled on the target host (as I understand it is nowadays default for OI151_a...) What are the suggestions to solve this? I tried several approaches with sudo, and su to no avail. I had tried to enable pfexec on the target system, too and couldn't do it. you can run it over any tcp transport. Do it yourself options include using ttcp or netcat as a transport, but almost anything will do. It requires a little bit more synchronization to do it this way since the receiver must be running (piped into zfs recv) before the transmitter is started or the transmitter will abort. In the end, though, you need to run the zfs recv as root somehow. If that's the crux of the problem and not the inability to ssh as root, you'll have to figure out a fix to get root at least for the zfs recv process. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS; what the manuals don't say ...
On 10/23/2012 8:29 AM, Robin Axelsson wrote: Hi, I've been using zfs for a while but still there are some questions that have remained unanswered even after reading the documentation so I thought I would ask them here. I have learned that zfs datasets can be expanded by adding vdevs. Say that you have created say a raidz3 pool named mypool with the command # zpool create mypool raidz3 disk1 disk2 disk3 ... disk8 you can expand the capacity by adding vdevs to it through the command # zpool add mypool raidz3 disk9 disk10 ... disk16 The vdev that is added doesn't need to have the same raid/mirror configuration or disk geometry, if I understand correctly. It will merely be dynamically concatenated with the old storage pool. The documentations says that it will be striped but it is not so clear what that means if data is already stored in the old vdevs of the pool. Unanswered questions: * What determines _where_ the data will be stored on a such a pool? Will it fill up the old vdev(s) before moving on to the new one or will the data be distributed evenly? * If the old pool is almost full, an even distribution will be impossible, unless zpool rearranges/relocates data upon adding the vdev. Is that what will happen upon adding a vdev? * Can the individual vdevs be read independently/separately? If say the newly added vdev faults, will the entire pool be unreadable or will I still be able to access the old data? What if I took a snapshot before adding the new vdev? * Can several datasets be mounted to the same mount point, i.e. can multiple file system-datasets be mounted so that they (the root of them) are all accessed from exactly the same (POSIX) path and subdirectories with coinciding names will be merged? The purpose of this would be to seamlessly expand storage capacity this way just like when adding vdevs to a pool. * If that's the case how will the data be distributed/allocated over the datasets if I copy a data file to that path? Kind regards Robin. *) yes, you can dynamically add more disks and zfs will just start using them. *) zfs stripes across all vdevs evenly, as it can. *) as your old vdev gets full, zfs will only allocate blocks to the newer, less full vdev *) since it's a stripe across vdevs (and they should all be raidz2 or better!) if one vdev fails, your filesystem will be unavailable. They are not independent unless you put them in a separate pool. *) you cannot have overlapping /mixed filesystems at exactly the same place, however it is perfectly possible to have e.g. /export be on rootpool, /export/mystuff on zpool1 and /export/mystuff/morestuff be on zpool2. The unasked question is If I wanted the vdevs to be equally balanced, could I?. The answers is a qualified yes. What you would need to do is reopen every single file, buffer it to memory, then write every block out again. We did this operation once. It means that all vdevs will roughly have the same block allocation when you are done. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS; what the manuals don't say ...
On 10/23/2012 11:08 AM, Robin Axelsson wrote: On 2012-10-23 15:41, Doug Hughes wrote: On 10/23/2012 8:29 AM, Robin Axelsson wrote: Hi, I've been using zfs for a while but still there are some questions that have remained unanswered even after reading the documentation so I thought I would ask them here. I have learned that zfs datasets can be expanded by adding vdevs. Say that you have created say a raidz3 pool named mypool with the command # zpool create mypool raidz3 disk1 disk2 disk3 ... disk8 you can expand the capacity by adding vdevs to it through the command # zpool add mypool raidz3 disk9 disk10 ... disk16 The vdev that is added doesn't need to have the same raid/mirror configuration or disk geometry, if I understand correctly. It will merely be dynamically concatenated with the old storage pool. The documentations says that it will be striped but it is not so clear what that means if data is already stored in the old vdevs of the pool. Unanswered questions: * What determines _where_ the data will be stored on a such a pool? Will it fill up the old vdev(s) before moving on to the new one or will the data be distributed evenly? * If the old pool is almost full, an even distribution will be impossible, unless zpool rearranges/relocates data upon adding the vdev. Is that what will happen upon adding a vdev? * Can the individual vdevs be read independently/separately? If say the newly added vdev faults, will the entire pool be unreadable or will I still be able to access the old data? What if I took a snapshot before adding the new vdev? * Can several datasets be mounted to the same mount point, i.e. can multiple file system-datasets be mounted so that they (the root of them) are all accessed from exactly the same (POSIX) path and subdirectories with coinciding names will be merged? The purpose of this would be to seamlessly expand storage capacity this way just like when adding vdevs to a pool. * If that's the case how will the data be distributed/allocated over the datasets if I copy a data file to that path? Kind regards Robin. *) yes, you can dynamically add more disks and zfs will just start using them. *) zfs stripes across all vdevs evenly, as it can. *) as your old vdev gets full, zfs will only allocate blocks to the newer, less full vdev *) since it's a stripe across vdevs (and they should all be raidz2 or better!) if one vdev fails, your filesystem will be unavailable. They are not independent unless you put them in a separate pool. *) you cannot have overlapping /mixed filesystems at exactly the same place, however it is perfectly possible to have e.g. /export be on rootpool, /export/mystuff on zpool1 and /export/mystuff/morestuff be on zpool2. The unasked question is If I wanted the vdevs to be equally balanced, could I?. The answers is a qualified yes. What you would need to do is reopen every single file, buffer it to memory, then write every block out again. We did this operation once. It means that all vdevs will roughly have the same block allocation when you are done. Do you happen to know how that's done in OI? Otherwise I would have to move each file one by one to a disk location outside the dataset and then move it back or zfs send the dataset to another pool of at least equal size and then zfs receive it back to the expanded pool. you don't have to move it, you just have to open, read it into memory, seek back to the beginning, and write it out again. Rewriting those blocks will take care of it since ZFS is copy-on-write. You will need to be wary of your snapshots during this process since all files will be rewritten and you'll double your space consumption. (basically a perl, python, or other similar script could do this) ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS remote receive
On 10/23/2012 4:13 PM, Timothy Coalson wrote: Works pretty well, though I get ~70MB/s on gigabit ethernet instead of the theoretically possible 120MB/s, and I'm not sure why (NFS gets pretty close to 120MB/s on the same network). There's a fair bit of overhead to ssh and to zfs send/recive, so you're doing very well. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Zfs stability
yes, you shoud do a scrub and no, there isn't very much risk to this. This will scan your disks for bits that have gone stale or the like. You should do it. We do a scrub once per week. On Fri, Oct 12, 2012 at 3:55 PM, Roel_D openindi...@out-side.nl wrote: Being on the list and reading all ZFS problem and question posts makes me a little scared. I have 4 Sun X4140 servers running in the field for 4 years now and they all have ZFS mirrors (2x HD). They are running Solaris 10 and 1 is running solaris 11. I also have some other servers running OI, also with ZFS. The Solaris servers N E V E R had any ZFS scrub. I didn't even knew such existed ;-) Since it all worked flawless for years now i am a huge Solaris/OI fan. But how stable are things nowaday? Does one need to do a scrub? Or a resilver? How come i see so much ZFS trouble? Kind regards, The out-side ___ 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] Zfs stability Scrubs
So?}?\, a lot of people have already answered this in various ways. I'm going to provide a little bit of direct answer and focus to some of those other answers (and emphasis) On 10/12/2012 5:07 PM, Michael Stapleton wrote: It is easy to understand that zfs srubs can be useful, But, How often do we scrub or the equivalent of any other file system? UFS? VXFS? NTFS? ... ZFS has scrubs as a feature, but is it a need? I do not think so. Other file systems accept the risk, mostly because they can not really do anything if there were errors. That's right. They cannot do anything. Why is that a good thing? If you have a corruption on your filesystem because a block or even a single bit went wrong, wouldn't you want to know? Wouldn't you want to fix it? What if a number in an important financial document changed? Seems unlikely, but we've discovered at least 5 instances of spontaneous disk data corruption over the course of a couple of years. zfs corrected them transparently. No data lost, automatic, clean, and transparent. The more data that we make, the more that possibility of spontaneous data corruption becomes reality. It does no harm to do periodic scrubs, but I would not recommend doing them often or even at all if scrubs get in the way of production. What is the real risk of not doing scrubs? data changing without you knowing it. Maybe this doesn't matter on an image file (though a jpeg could end up looking nasty or destroyed, and mpeg4 could be permanently damaged, but in a TIFF or other uncompressed format, you'd probably never know) Risk can not be eliminated, and we have to accept some risk. For example, data deduplication uses digests on data to detect duplication. Most dedup systems assume that if the digest is the same for two pieces of data, then the data must be the same. This assumption is not actually true. Two differing pieces of data can have the same digest, but the chance of this happening is so low that the risk is accepted. but, the risk of data being flipped once you have TBs of data is way above 0%. You can also do your own erasure coding if you like. That would be one way to achieve the same affect outside of ZFS. I'm only writing this because I get the feeling some people think scrubs are a need. Maybe people associate doing scrubs with something like doing NTFS defrags? NTFS defrag would only help with performance. scrub helps with integrity. Totally different things. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] rebuilding man page catalog
catman -w rebuilds the index (-M to supply directory) On Tue, Oct 9, 2012 at 11:18 AM, Boris Epstein borepst...@gmail.com wrote: Hello listmates, If may man command does not display some of the pages including those clearly present under /usr/share/man - how do I fix that? I remember there was a command that rebuilt the catalog, I just can't recall what it was. Thanks. Boris. ___ 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] How to test ethernet throughput?
A third recommendation for iperf. It's the tool you want. Don't mess around with anything else. On 5/18/2012 10:02 AM, Rich wrote: iperf is also your friend, as is dd+netcat+pv if you want something a bit less rigorous. - Rich ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] nfs mounting and cascading dataset access?
On 2/27/2012 11:32 AM, James Carlson wrote: Scott LeFevre wrote: I have a pools setup named tank1 with multiple datasets/filesystems defined. At the top, I have tank1/media followed by tank1/media/Video and tank1/media/Music. I've setup tank1/media as a nfs share (e.g. /export/media) and can mount it from other systems. When I try to access media/Video from the nfs client, it shows an empty directory. (From the oi 151a2 host, everything shows up locally.) I can mount tank1/media/Video directly from the nfs client and everything shows up as it should. Is there a way to make this work where I expose just tank1/media as an nfs export and can access the underlying datasets/filesystems (e.g. Video and Music)? This isn't really a special issue with ZFS. NFS has always done exports by file system. If you have /foo and /foo/bar, and /foo/bar is a mount point for some other file system, then exporting /foo does not give clients access to the contents of /foo/bar, because there's no way to cross the file system boundary. Instead, clients looking at /foo/bar will see there whatever is actually in the /foo file system -- probably just an underlying empty directory used as the mount point. What you can do is set sharenfs=on (or some list of NFS options instead of just on) using the zfs command on tank/media, and all of the filesystems under that point will inherit this setting. That means that clients will see separate NFS exports for each of the underlying datasets -- one for Video and one for Music in this case. Automount, of course, is your friend. ;-} You can do this (hierarchical implicit shared subfs) with NFSv4, but not with NFSv3. There's a lot more setup for V4 though and it would likely not be as robust. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ZFS send recv soln
Yes, that's a very fine way to do it and we do it here. I'm afraid our python scripts wouldn't be much use to you though. They are very site specific. My main point for posting is a word of encouragement. On Mon, Jan 9, 2012 at 5:47 PM, Marcus Dillury mar...@dillury.net wrote: Hi, I have a customer who would like to use a zfs backup solution for his production zpool. I was thinking of using incremental snapshot send/recv to do this. I was wondering if anyone has any useful scripts that can facilitate the recursive send/recv of approx 2000 zfs datasets. My client has limited technical experience, so I was ideally after something that can automatically remove old snapshots when space becomes constrained. Thanks in advance. -Marcus ___ 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] Fileserver performance with log and cache devices?
On 12/1/2011 3:15 PM, David Brodbeck wrote: On Thu, Dec 1, 2011 at 8:09 AM, Geoff Flaritygeoff.flar...@gmail.comwrote: My advice would be to max out your RAM (for ARC) before you bother with a ZIL, or L2ARC. Where a fast SSD for a ZIL really shines is random synchronized writes. IE a database transaction. You'll notice this when you run filebench and look at the results. I'm not sure how NFS handles FSYNC and DSYNC to be honest. NFS does a *lot* of synchronous writes and will probably benefit from a fast ZIL device. It's particularly valuable data for heavy meta-data intensive things with NFS, which includes deleting directories with rm -r and creating or updating lots of small files. NFS commits beat on the ZIL. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Supermicro SC216 chassis with multiple HBAs?
On 10/26/2011 7:40 PM, Jason J. W. Williams wrote: Has anyone here tried using two LSI HBAs (9211-8i) in a single SC216 chassis? I'm being told there's only one iPass cable off the backplane and so only one HBA can be used...but with the number of folks adding in SATA SSDs it seems like there must be a solution so you don't have to mix SATA and SAS on the same HBA. Any insight is greatly appreciated. My integrator is configuring with 3 of the 9211-8i..Either they know something, or they don't know something.. But I have 3 integrators essentially offering the samething, so I assume they all must know something. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] ssh sessions timing out?
On 10/12/2011 10:56 PM, Dan Swartzendruber wrote: This one is driving me nuts. I can't seem to keep an ssh session open for more than 10 minutes or so. When I check after that, putty has an error up about network error caused software abort (or words to that effect.) I have TCP/IP keepalives turned on, as well as ssh keepalives, and that hasn't helped. This is the only server that has this problem :( I've double and tripled checked the putty settings and don't see anything helpful. Any ideas? Thanks! you could try running sshd -d, or put a 'truss' onto the sshd after running to see what it's doing. ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] Help with website
On 10/11/2011 6:10 AM, Dmitry Kozhinov wrote: doesn't pull let them manage their own time better? Yes. This is why my vote is for forum, not mailing list. Dmitry. I've stayed out of this thread thus far, and probably will going forward, but I vastly prefer mail lists and I think the above is a non-argument. Forums do not allow better time management by any objective means. In a mail list, I am equally capable of sending all of the mail list to a folder, or not, threaded in my mail reader, or not, as I choose, and I can read them whenever I want. If I see a topic pop up that interests me, I can read it then, or not. I can make agile changes in my interests easily. Forums have one major failing for me: Out of sight is out of mind. If I have to go to a website, login (maybe it has cookies, maybe it doesn't; maybe it does and I don't want the website to remember my password!) then find the thread or topic where I left off, which will take a few clicks or more depending upon how long it has been. Mail messages are all right there, and I only need to go to the most recent undeleted or unread one with less effort. But, there is also a tradeoff in volume. If a mail list starts receiving 100s of messages a day it becomes less useful and it might as well be a forum. Linux kernel development lists come to mind. Those would be more useful as a forum. There are very few topics in there that actually might interest me and the flood of mail makes it a chore. OI is a smallish mail list (aside from this thread!) so managing the (normally) 20 messages that come in in a day is easier in my inbox. thresholds matter. $.02 ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
[OpenIndiana-discuss] OI vs U9 performance surprises
Box = Sun x4240 with 8xIntel 320 160GB flash R6x6 SSD oi R6x6 SSD u9 R5x5 SSD u9 R5x6u9 R6 ssd oi/12 100k4m 37s 3m 31s 3m 24s 3m 27s 4m 18s rm 100k 1m 39s 1m 3s 1m 1m 1m 25s star-x 3m 5s 3m 37s 2m 37s 2m 43s 2m 53s cp b.tar2m 14s 2m 18s 2m 8s 2m 9s 2m 1s dual touch 3m 41s 3m 35s 3m 43s 4m 24s dual rm 1m 4s 1m 2s 1m 4s 1m 25s dual star -x2.5/5m 3.3m 4.5m 3m/4m 3.5m 4.5m d cp 12.7G 3m 45s 3m 45s 4m 3m 45s SSD = Intel 320 160GB drives R6 = 6 drives in raid6 config u9 = Solaris 10 update 9 R5x5 = Intel 320 160GB in raid5 with 5 disks oi= OpenIndiana 148 release oi/12 = OpenIndiana 148 release with zpool-12 patch for ashift=12 100k = 100k files: 1000 directories, 100 files each, using touch (0 byte files - metadata/ops intensive) rm 100k = removing all files and directories star -x = extracting (over NFS, 1gig) a 12.7GB file with a mix of small and large file cp b.tar = copying 12.7GB tar file over NFS to server (bounded by wire) dual touch = touch 100k files over NFS on dedicated 1gig links to server dual rm = removing those files dual cp = copying the 12.7GB tar file from the node (bounded by 1gbit wire again) The odd thing is that, aside from the things that are bounded by 1gbit wire, that u9 is quite a bit faster than OpenIndiana. I have 128 NFS servers defined, and they never seem to be bottlenecked. I would post the numbers for 73GB spinning media, but the run times for OpenIndiana were unboundedly slow, and I hadn't been able to figure that out (hours later, I had to ctrl-c vs. minutes on U9). when I ran dstrace ops script against the 73GB disks it was taking hundreds of mseconds for ops to complete. But, since I haven't posted the performance numbers of the 73GB disks here (will re-do benchmarks in a bit and repost if asked), I guess we could focus on why OI appears to be significantly slower. ashift=12 improves things by a small amount, but still not to the point of u9 speed. Advice/tips welcome. This is a relatively new/virgin OI-148 install. (one note, the double numbers often include 2 numbers. It appears that the using dladm on the second interface to create an aggregate has a signficant impact on transfer speed/latency, but we can ignore that for now) ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] OI vs U9 performance surprises
On 5/9/2011 12:01 PM, Doug Hughes wrote: Box = Sun x4240 with 8xIntel 320 160GB flash R6x6 SSD oi R6x6 SSD u9 R5x5 SSD u9 R5x6u9 R6 ssd oi/12 100k4m 37s 3m 31s 3m 24s 3m 27s 4m 18s rm 100k 1m 39s 1m 3s 1m 1m 1m 25s star-x 3m 5s 3m 37s 2m 37s 2m 43s 2m 53s cp b.tar2m 14s 2m 18s 2m 8s 2m 9s 2m 1s dual touch 3m 41s 3m 35s 3m 43s 4m 24s dual rm 1m 4s 1m 2s 1m 4s 1m 25s dual star -x2.5/5m 3.3m 4.5m 3m/4m 3.5m 4.5m d cp 12.7G 3m 45s 3m 45s 4m 3m 45s SSD = Intel 320 160GB drives R6 = 6 drives in raid6 config u9 = Solaris 10 update 9 R5x5 = Intel 320 160GB in raid5 with 5 disks oi= OpenIndiana 148 release oi/12 = OpenIndiana 148 release with zpool-12 patch for ashift=12 100k = 100k files: 1000 directories, 100 files each, using touch (0 byte files - metadata/ops intensive) rm 100k = removing all files and directories star -x = extracting (over NFS, 1gig) a 12.7GB file with a mix of small and large file cp b.tar = copying 12.7GB tar file over NFS to server (bounded by wire) dual touch = touch 100k files over NFS on dedicated 1gig links to server dual rm = removing those files dual cp = copying the 12.7GB tar file from the node (bounded by 1gbit wire again) The odd thing is that, aside from the things that are bounded by 1gbit wire, that u9 is quite a bit faster than OpenIndiana. I have 128 NFS servers defined, and they never seem to be bottlenecked. I would post the numbers for 73GB spinning media, but the run times for OpenIndiana were unboundedly slow, and I hadn't been able to figure that out (hours later, I had to ctrl-c vs. minutes on U9). when I ran dstrace ops script against the 73GB disks it was taking hundreds of mseconds for ops to complete. But, since I haven't posted the performance numbers of the 73GB disks here (will re-do benchmarks in a bit and repost if asked), I guess we could focus on why OI appears to be significantly slower. ashift=12 improves things by a small amount, but still not to the point of u9 speed. Advice/tips welcome. This is a relatively new/virgin OI-148 install. (one note, the double numbers often include 2 numbers. It appears that the using dladm on the second interface to create an aggregate has a signficant impact on transfer speed/latency, but we can ignore that for now) Here are some dtrace ops numbers for the server side (linux client, same client for both) with open indiana: hughesd@solaris:/var/tmp/mravi# bin/nfs-ops-response.d dtrace: script 'bin/nfs-ops-response.d' matched 3 probes ^C CPU FUNCTION 1 | :END NFS v3 Ops count == RFS3_FSSTAT 2 RFS3_MKDIR 1000 RFS3_GETATTR 1001 RFS3_ACCESS1006 RFS3_CREATE 101000 RFS3_SETATTR 101000 RFS3_LOOKUP 101003 NFS v3 Ops average RESPONSE time (usec) RFS3_GETATTR 39 RFS3_LOOKUP 46 RFS3_ACCESS 52 RFS3_FSSTAT 304 RFS3_SETATTR449 RFS3_CREATE 508 RFS3_MKDIR 667 NFS v3 Ops average SYSTEM time (usec) == RFS3_GETATTR 32 RFS3_ACCESS 40 RFS3_LOOKUP 40 RFS3_FSSTAT 80 RFS3_SETATTR143 RFS3_CREATE 200 RFS3_MKDIR 340 --- here are numbers with Sol 10 U9 (same server hardware, same client) 6 disk raidz1: dtrace: script '/var/tmp/mravi/bin/nfs-ops-response.d' matched 3 probes ^C CPU FUNCTION 5 | :END NFS v3 Ops count == RFS3_GETATTR 1000 RFS3_MKDIR 1000 RFS3_ACCESS1004 RFS3_CREATE 101000 RFS3_LOOKUP 101000 RFS3_SETATTR 101000 NFS v3 Ops average RESPONSE time (usec) RFS3_LOOKUP 17 RFS3_GETATTR 18 RFS3_ACCESS 21 RFS3_SETATTR270 RFS3_CREATE 294 RFS3_MKDIR 340 NFS v3 Ops average SYSTEM time (usec) == RFS3_GETATTR 16 RFS3_LOOKUP 16 RFS3_ACCESS 18 RFS3_SETATTR 54 RFS3_CREATE 75 RFS3_MKDIR 124 --- Here's the dtrace script #!/usr/sbin/dtrace -FCs /* # Author: Ravi Mallarapu # Date: 20101117 # Description: display NFS v3 ops count and average user and system response time # */ fbt:nfssrv:common_dispatch:entry /args[0]-rq_vers == 3 ! self-trace
[OpenIndiana-discuss] dladm create-aggr fails with link busy
fresh openinidna 148 install. It came up with ipv6 for nge1/nge2 after auto-discovering and I did the unplumb on those, so the interfaces are clean and unused, but still I can't create the aggregate. hughesd@x4240-3-1-17:~# dladm show-aggr hughesd@x4240-3-1-17:~# dladm create-aggr -d nge2 1 dladm: create operation failed: link busy hughesd@x4240-3-1-17:~# dladm create-aggr -d nge1 1 dladm: create operation failed: link busy hughesd@x4240-3-1-17:~# dladm show-aggr hughesd@x4240-3-1-17:~# ifconfig -a lo0: flags=2001000849UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL mtu 8232 index inet 127.0.0.1 netmask ff00 nge0: flags=1004843UP,BROADCAST,RUNNING,MULTICAST,DHCP,IPv4 mtu 1500 index 8 inet 10.232.4.60 netmask fc00 broadcast 10.232.7.255 ether 0:14:4f:ed:b8:ec hughesd@x4240-3-1-17:~# uname -a SunOS x4240-3-1-17 5.11 oi_148 i86pc i386 i86pc hughesd@x4240-3-1-17:~# dladm create-aggr -d nge1 -d nge2 1 dladm: create operation failed: link busy hughesd@x4240-3-1-17:~# ___ OpenIndiana-discuss mailing list OpenIndiana-discuss@openindiana.org http://openindiana.org/mailman/listinfo/openindiana-discuss
Re: [OpenIndiana-discuss] dladm create-aggr fails with link busy
that did it! thanks! (never would have found that) On 5/6/2011 12:25 PM, Lucas Van Tol wrote: I saw something like this when I hadn't disabled NWAM before trying to create an aggregate... Date: Fri, 6 May 2011 11:43:39 -0400 From: d...@will.to To: openindiana-discuss@openindiana.org Subject: [OpenIndiana-discuss] dladm create-aggr fails with link busy fresh openinidna 148 install. It came up with ipv6 for nge1/nge2 after auto-discovering and I did the unplumb on those, so the interfaces are clean and unused, but still I can't create the aggregate. hughesd@x4240-3-1-17:~# dladm show-aggr hughesd@x4240-3-1-17:~# dladm create-aggr -d nge2 1 dladm: create operation failed: link busy hughesd@x4240-3-1-17:~# dladm create-aggr -d nge1 1 dladm: create operation failed: link busy hughesd@x4240-3-1-17:~# dladm show-aggr hughesd@x4240-3-1-17:~# ifconfig -a lo0: flags=2001000849UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL mtu 8232 index inet 127.0.0.1 netmask ff00 nge0: flags=1004843UP,BROADCAST,RUNNING,MULTICAST,DHCP,IPv4 mtu 1500 index 8 inet 10.232.4.60 netmask fc00 broadcast 10.232.7.255 ether 0:14:4f:ed:b8:ec hughesd@x4240-3-1-17:~# uname -a SunOS x4240-3-1-17 5.11 oi_148 i86pc i386 i86pc hughesd@x4240-3-1-17:~# dladm create-aggr -d nge1 -d nge2 1 dladm: create operation failed: link busy hughesd@x4240-3-1-17:~# ___ 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