Re: [SCIENTIFIC-LINUX-USERS] firefox 45.1 crashes
On 29/04/16 17:43, Graham Allan wrote: > On 4/29/2016 10:37 AM, Iosif Fettich wrote: >> >> I'll just second Pat: no problem encountered so far, despite heavy use. >> >> Just in case that might make a difference (it did, occasionally, in the >> past): I'm having 16 GiB of memory in my desktop. >> >> Iosif Fettich > > Good point; I have a machine with 16GB or so, and no problems, but we > certainly have some people with old/weak machine - we'll check if the amount > of RAM correlates at all. I've run Firefox 45 ESR since the release in RHEL7.2 on a machine with 8GB RAM without any issues, and with 10-15 tabs open. Also upgraded a my laptops with SL7.2, but too early to say much about the stability. But I haven't really noticed anything special. -- kind regards, David Sommerseth
Re: SL7.2 Live DVDkde would not boot
On 03/04/16 17:10, John Pilkington wrote: >> > > I've had an interesting week with a new 3TB drive and a family box that has > been running MS Vista for years. I disconnected the Windows HD 'for safety' > and installed kubuntu from the live DVD, with few problems until I tried a > 'real' boot, which failed. Eventually I installed buntu 14 with grub > alongside Vista on the original HD, and also have buntu 16 beta on the new > one; at present they will all boot and run. Don't know if SL7 would do the > same. But the USB drive exploit looks handy. This should work on the majority of all Linux distributions, at least if you use UUID for the /boot partitionsi. Use of LVM can also simplify mounting the root parition (/) and so on - unless you use UUID for those mount points too. I've booted several old Linux installations from hardrives put into a USB closure. I haven't tried to do that with Windows though, that might work too - but somehow I imagine it will freak out at some point where drive letters won't match properly. To get an overview you can run 'blkid' or 'lsblk -o NAME,UUID' on your system to see all devices and their unique UUID. These tools are also valuable when you need to modify /etc/crypttab manually. -- kind regards, David Sommerseth
Re: How does NetworkManager monitor the connection files?
On 31/03/16 00:10, Benjamin Lefoul wrote: > But sed -i ALSO changes the inode, and as I said it doesn't work: > > root@hoptop:~# touch a > root@hoptop:~# ls -i a > 9700011 a > root@hoptop:~# sed -i 's/q/a/g' a > root@hoptop:~# ls -i a > 9700013 a > I've not looked into the NM code. But it wouldn't surprise me that much if inotify is used. But it might be it doesn't catch all the events which sed would trigger, but only modification triggered by editors. Can you also check what happens to the inode on a file if you use vim/emacs? -- kind regards, David Sommerseth > > From: owner-scientific-linux-us...@listserv.fnal.gov > <owner-scientific-linux-us...@listserv.fnal.gov> on behalf of Tom H > <tomh0...@gmail.com> > Sent: 30 March 2016 23:00 > To: SL Users > Subject: Re: How does NetworkManager monitor the connection files? > > On Wed, Mar 30, 2016 at 3:49 PM, Benjamin Lefoul > <benjamin.lef...@nwise.se> wrote: >> >> I have set monitor-connection-files=true in my >> /etc/NetworkManager/NetworkManager.conf >> >> It works fine (in fact, instantly) if I edit >> /etc/sysconfig/network-scripts/ifcfg-eth0 with emacs or vi (for instance, >> changing the IP). >> >> It fails miserably if I use sudoedit, or sed: >> >> # grep 100 /etc/sysconfig/network-scripts/ifcfg-eth0 >> IPADDR=192.168.4.100 >> >> # sed -i 's/100/155/g' /etc/sysconfig/network-scripts/ifcfg-eth0 >> >> Even though all stats (access modify and change) are renewed. >> >> It's worse than that: even nmcli con reload afterwards fails. >> >> In fact, the only way to get the ip to change is by entering the file with >> vi, not touching it, and leave with ":wq" (not just ":q"). >> >> Why is that? What is going on here? >> >> I know, I know, I can use nmcli in scripts, and not string-manipulation >> tools, but say I don't want to... :) >> >> And still, during operations, I'd rather edit the files with sudoedit... > > "sudo -e ifcfg-file" doesn't change the inode. Can you use "sudo vi > ifcfg-file"? (Or whichever editor you prefer.) >
Re: named-chroot issue - AGAIN
On 17/03/16 12:40, Steven Haigh wrote: > On 17/03/2016 10:25 PM, David Sommerseth wrote: >> On 17/03/16 06:36, Bill Maidment wrote: >>> Hi guys >>> Another named update and still the named-chroot.service file has not been >>> fixed. It is really annoying to have to manually fix it every time, just to >>> get DNS working after an update. >>> Why is the -t /var/named/chroot option included in the ExecStart but not in >>> the ExecStartPre >>> >>> ExecStartPre=/bin/bash -c 'if [ ! "$DISABLE_ZONE_CHECKING" == "yes" ]; then >>> /usr/sbin/named-checkconf -z /etc/named.conf; else echo "Checking of zone >>> files is disabled"; fi' >>> ExecStart=/usr/sbin/named -u named -t /var/named/chroot $OPTIONS >>> >>> Surely named-checkconf should be run with the same named.conf file as named >>> !!! >>> >>> This was reported back in November 2015 >>> https://bugzilla.redhat.com/show_bug.cgi?id=1278082 >>> This should have been fixed by now. How hard is a one line change to fix ??? >> >> This bug has severity set to medium. That means it has most likely not been >> considered critical enough by Red Hat to go into an errata in the 7.2 life >> cycle. But as the status is ASSIGNED (not NEW, which is the first status >> level) - it means someone is working on it. >> >> If you do not like this pace, you can log in to the Red Hat customer portal >> and get in touch with Red Hat support. If you can provide them with good >> technical arguments why this must be added in the 7.2 life cycle, then you >> might see this fixed sooner. >> >> If you do not have Red Hat subscription with support ... well, then you need >> to patiently wait. Scientific Linux builds on the source RPMs Red Hat >> releases. >> >> And the reason for these things to take time is that every BZ for RHEL goes >> through several steps of quality control before a fix gets released. It means >> Red Hat needs to allocate resources getting these bugs fixed, verified and >> tested before users see the update. This is the key concept of enterprise >> distributions, to put efforts into avoiding regressions or new bugs as much >> as >> possible and to try to deliver a stable distribution which is properly >> maintained and updated over many years. > > I would agree with you if they didn't remove that option in a release. > the -t /blah was actually removed in a commit - which QA failed to pick > up (and likely chroot bind setup wasn't tested at all). > > I don't think this is a great example of what RedHat does well - this is > an example of what they do *really* bad. Not going to argue that this could have been done better, I agree with you here. On the other hand, maybe *that* is one reason it takes time to get this issue resolved too? That Red Hat QE is working on improving the situation, adding needed regression tests and so on for this use case. I know I'm speculating now, but I also know that these guys really do their best to avoid painful experiences for users and customers. Unfortunately, they do mistakes - as we all do from time to time. -- kind regards, David Sommerseth
Re: named-chroot issue - AGAIN
On 17/03/16 13:23, Tom H wrote: > On Thu, Mar 17, 2016 at 12:53 PM, David Sommerseth > <sl+us...@lists.topphemmelig.net> wrote: >> >> Not going to argue that this could have been done better, I agree with you >> here. On the other hand, maybe *that* is one reason it takes time to get this >> issue resolved too? That Red Hat QE is working on improving the situation, >> adding needed regression tests and so on for this use case. I know I'm >> speculating now, but I also know that these guys really do their best to >> avoid >> painful experiences for users and customers. Unfortunately, they do mistakes >> - as we all do from time to time. > > Given the > https://git.centos.org/blobdiff/rpms!bind.git/d56ed2d3a2736a07a09c268f3b2607cca8f1b6ca/SOURCES!named-chroot.service > commit, there's probably a lot of hype in RH's QA marketing claims. > I'm not implying that there's no QA at all but, in this case, if there > was any, it sucked. The people working on CentOS are not the same people working on RHEL, even if they're working in the same company. And RHEL is still the upstream source of CentOS. So if CentOS decides to fix this on their own, they need to keep track of this until it's fixed in RHEL and then remove their own fix. Of course SL could do that as well, but that can be a maintenance burden. That's the downside of being a downstream distro. -- kind regards, David Sommerseth
Re: named-chroot issue - AGAIN
On 17/03/16 06:36, Bill Maidment wrote: > Hi guys > Another named update and still the named-chroot.service file has not been > fixed. It is really annoying to have to manually fix it every time, just to > get DNS working after an update. > Why is the -t /var/named/chroot option included in the ExecStart but not in > the ExecStartPre > > ExecStartPre=/bin/bash -c 'if [ ! "$DISABLE_ZONE_CHECKING" == "yes" ]; then > /usr/sbin/named-checkconf -z /etc/named.conf; else echo "Checking of zone > files is disabled"; fi' > ExecStart=/usr/sbin/named -u named -t /var/named/chroot $OPTIONS > > Surely named-checkconf should be run with the same named.conf file as named > !!! > > This was reported back in November 2015 > https://bugzilla.redhat.com/show_bug.cgi?id=1278082 > This should have been fixed by now. How hard is a one line change to fix ??? This bug has severity set to medium. That means it has most likely not been considered critical enough by Red Hat to go into an errata in the 7.2 life cycle. But as the status is ASSIGNED (not NEW, which is the first status level) - it means someone is working on it. If you do not like this pace, you can log in to the Red Hat customer portal and get in touch with Red Hat support. If you can provide them with good technical arguments why this must be added in the 7.2 life cycle, then you might see this fixed sooner. If you do not have Red Hat subscription with support ... well, then you need to patiently wait. Scientific Linux builds on the source RPMs Red Hat releases. And the reason for these things to take time is that every BZ for RHEL goes through several steps of quality control before a fix gets released. It means Red Hat needs to allocate resources getting these bugs fixed, verified and tested before users see the update. This is the key concept of enterprise distributions, to put efforts into avoiding regressions or new bugs as much as possible and to try to deliver a stable distribution which is properly maintained and updated over many years. -- kind regards, David Sommerseth
Re: "Not using downloaded repomd.xml because it is older than what we have:"
On 9 March 2016 06:05:04 CET, Eero Volotinen <eero.voloti...@iki.fi> wrote: >Do you have transparent/forced proxy in network? It happens for me also on boxes connected directly to the internet; no proxy involved. Every single day I see this this message. -- kind regards, David Sommerseth >ke 9. maaliskuuta 2016 klo 6.53 Thomas Leavitt <tleav...@eag.com> >kirjoitti: > >> No feedback? Is everyone else just ignoring these messages? >> >> Thomas >> >> -Original Message- >> From: owner-scientific-linux-us...@listserv.fnal.gov [mailto: >> owner-scientific-linux-us...@listserv.fnal.gov] On Behalf Of Thomas >> Leavitt >> Sent: Monday, February 29, 2016 10:53 AM >> To: scientific-linux-us...@fnal.gov >> Subject: RE: "Not using downloaded repomd.xml because it is older >than >> what we have:" >> >> Not using downloaded repomd.xml because it is older than what we >have: >> Current : Tue Feb 16 08:58:20 2016 >> Downloaded: Tue Feb 16 08:58:13 2016 >> >> A 7 second variance shouldn't be that much of an issue. >> >> Regards, >> Thomas Leavitt >> >> -Original Message- >> From: Thomas Leavitt >> Sent: Monday, February 29, 2016 10:52 AM >> To: 'scientific-linux-us...@fnal.gov' >> Subject: RE: "Not using downloaded repomd.xml because it is older >than >> what we have:" >> >> I've been meaning to write about this for a while... my inbox is >flooded >> every day with messages with this as the content from my SL7 machines >(the >> ones I have set up to forward mail sent to root). I checked the time >on >> them, they're using NTP, and the time agrees, almost to the second, >with >> that of network time services and other machines, I doubt there's >even a 30 >> second variance. >> >> What's the strategy for dealing with this? Seems like it isn't an >isolated >> problem. >> >> Regards, >> Thomas Leavitt >> >> -Original Message- >> From: owner-scientific-linux-us...@listserv.fnal.gov [mailto: >> owner-scientific-linux-us...@listserv.fnal.gov] On Behalf Of David >> Sommerseth >> Sent: Thursday, February 18, 2016 2:51 AM >> To: Peter Boy; SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV >> Subject: Re: "Not using downloaded repomd.xml because it is older >than >> what we have:" >> >> On 13/02/16 19:24, Peter Boy wrote: >> > Hi all, >> > >> > since several months I get constantly from anacreon: >> >> ——< >> > /etc/cron.daily/0yum-daily.cron: >> > >> > Not using downloaded repomd.xml because it is older than what we >have: >> > Current : Thu Feb 4 16:13:26 2016 >> > Downloaded: Thu Feb 4 16:13:25 2016 >> >> ——< >> > >> > >> > The time difference is quite minimal. And a manual „yum update“ >> > confirms that no updates are waiting. >> > >> > Using my favourite search engine I found it might have be caused by >an >> > unresponsive of lazy mirror. But I use the standard configuration, >i.e. >> > the mirror list just includes the three scientificlinux servers. >Other >> > entries refer old bugs long fixed. >> > >> > I tried a yum clean all but it didn’t fix it. >> > >> > And all our other don’t show this issue, but the configuration is >all >> > the same, at least according to my knowledge. >> > >> > >> > Obviously, there is no harm done and it can be safely ignored. But >it >> > always pulls our issue alert button. >> >> Hi, >> >> I am seeing exactly the same. I thought it was NTP issues related to >my >> own setup, where I have a local rsync mirror. But then I installed >from >> scratch SL7 on another site without any local mirrors, and the same >issue >> appears there too. So I see this both with public repositories as >well as >> local rsync repositories. >> >> I have also seen this on SL6, but not as frequent as on SL7. >> >> Even though they cause no obvious harm, it gets quite annoying when >you >> receive many of them during a day ... sometimes even several days in >a row. >> >> Perhaps yum should be more graceful to the timestamp? And just quiet >> these messages if the time difference is less than 30 seconds or so. >> >> >> -- >> kind regards, >> >> David Sommerseth >> >> -- >> kind regards, >> >> David Sommerseth >> >> >> This e-mail may contain privileged or confidential information. If >you are >> not the intended recipient: (1) you may not disclose, use, >distribute, copy >> or rely upon this message or attachment(s); and (2) please notify the >> sender by reply e-mail, and then delete this message and its >attachment(s). >> EAG, Inc. and its affiliates disclaim all liability for any errors, >> omissions, corruption or virus in this message or any attachments. >> >>
Re: snooping windows 10 - how to stop it on a linux gateway?
On 05/03/16 13:23, David Sommerseth wrote: > On 05/03/16 11:36, jdow wrote: >> If squid can find usefully unique patterns in encrypted traffic I suppose >> that >> might work. But that's one heck of a big "if". > > A quick google search on "transparent https proxy" gave me these: > > <http://docs.mitmproxy.org/en/stable/howmitmproxy.html> > <http://rahulpahade.com/content/squid-transparent-proxy-over-ssl-https> > > I probably have more "faith" in the mitmproxy approach, as that seems > generally more designed with https in mind. Just another idea came to mind. You only need a transparent proxy to be used when connecting to IP ranges belonging to Microsoft. So instead of an iptables REDIRECT for all http/https connection, you add separate rules with --destination to the different Microsoft subnets. -- kind regards, David Sommerseth >> On 2016-03-05 02:15, Karel Lang AFD wrote: >>> Hmm ... yes, yes. >>> Thanks for bringing this up. >>> I force all http traffic through the squid proxy on our SL 6 gateway, this >>> could >>> be also helpful.. >>> >>> >>> >>> On 03/05/2016 11:00 AM, prmari...@gmail.com wrote: >>>> The only way I can think of is to force all internet access through a proxy >>>> and filter it out in the proxy. >>>> Then you don't give the machines any internet access just access to the >>>> proxy. >>>> Unfortunately I do not have details for you on how to filter the snoop >>>> messages because in I haven't looked at them but it should be fairly easy >>>> using squid and an external Perl regex filter script or other filter >>>> application, but you will take a latency hit because you will have to >>>> inspect >>>> every transaction. >>>> >>>>Original Message >>>> From: jdow >>>> Sent: Friday, March 4, 2016 23:35 >>>> To: scientific-linux-us...@fnal.gov >>>> Subject: Re: snooping windows 10 - how to stop it on a linux gateway? >>>> >>>> That windows update server is a relay for the "snoop" messages. About the >>>> only >>>> way to totally stop the snoop messages is to totally isolate the network >>>> containing Windows machines from the network. Any windows machine can serve >>>> as a >>>> relay point for any others. >>>> >>>> {o.o} >>>> >>>> On 2016-03-04 20:16, Karel Lang AFD wrote: >>>>> Hi guys, >>>>> >>>>> firstly, sorry Todd, i don't know how it happened i got attached to your >>>>> thread. >>>>> >>>>> secondly, thank you all for your thoughtful posts. >>>>> >>>>> I know it is not easy to block the selected traffic from windows 10 and >>>>> you are >>>>> right, it is being backported to windows 7 as well. Horrible and >>>>> disgusting. >>>>> >>>>> I already have windows server in LAN dedicated as a update server (work >>>>> of my >>>>> windows colleagues), so the PC don't have to access windows update servers >>>>> outside LAN - this should simplify things. >>>>> >>>>> Also the PCs must have internet access to email, http, https, ftp, sftp - >>>>> simply >>>>> the 'usual' stuff. >>>>> I think, yet, there should be a way. I'll try to consult mikrotik experts >>>>> (the >>>>> router brand we use) and guys from our ISP. >>>>> If i have something, i'll let you know :-) >>>>> >>>>> thank you, bb >>>>> >>>>> Karel >>>>> >>>>> On 03/05/2016 12:40 AM, Steven Haigh wrote: >>>>>> On 05/03/16 07:24, Karel Lang AFD wrote: >>>>>>> Hi all, >>>>>>> >>>>>>> guys, i think everyone heard already about how windows 10 badly treat >>>>>>> its users privacy. >>>>>> >>>>>> My solution to this was to finally rid Windows 7 off my desktop PC - as >>>>>> most of the telemetry has also been 'back ported' to Windows 7 also. You >>>>>> can't stop it. >>>>>> >>>>>>> I'm now thinking about a way howto stop a windows 10 sending these data >>>>>>> mining results to a microsoft telemetry servers and filter it on our SL >>>&g
Re: snooping windows 10 - how to stop it on a linux gateway?
On 05/03/16 11:36, jdow wrote: > If squid can find usefully unique patterns in encrypted traffic I suppose that > might work. But that's one heck of a big "if". A quick google search on "transparent https proxy" gave me these: <http://docs.mitmproxy.org/en/stable/howmitmproxy.html> <http://rahulpahade.com/content/squid-transparent-proxy-over-ssl-https> I probably have more "faith" in the mitmproxy approach, as that seems generally more designed with https in mind. -- kind regards, David Sommerseth > On 2016-03-05 02:15, Karel Lang AFD wrote: >> Hmm ... yes, yes. >> Thanks for bringing this up. >> I force all http traffic through the squid proxy on our SL 6 gateway, this >> could >> be also helpful.. >> >> >> >> On 03/05/2016 11:00 AM, prmari...@gmail.com wrote: >>> The only way I can think of is to force all internet access through a proxy >>> and filter it out in the proxy. >>> Then you don't give the machines any internet access just access to the >>> proxy. >>> Unfortunately I do not have details for you on how to filter the snoop >>> messages because in I haven't looked at them but it should be fairly easy >>> using squid and an external Perl regex filter script or other filter >>> application, but you will take a latency hit because you will have to >>> inspect >>> every transaction. >>> >>>Original Message >>> From: jdow >>> Sent: Friday, March 4, 2016 23:35 >>> To: scientific-linux-us...@fnal.gov >>> Subject: Re: snooping windows 10 - how to stop it on a linux gateway? >>> >>> That windows update server is a relay for the "snoop" messages. About the >>> only >>> way to totally stop the snoop messages is to totally isolate the network >>> containing Windows machines from the network. Any windows machine can serve >>> as a >>> relay point for any others. >>> >>> {o.o} >>> >>> On 2016-03-04 20:16, Karel Lang AFD wrote: >>>> Hi guys, >>>> >>>> firstly, sorry Todd, i don't know how it happened i got attached to your >>>> thread. >>>> >>>> secondly, thank you all for your thoughtful posts. >>>> >>>> I know it is not easy to block the selected traffic from windows 10 and >>>> you are >>>> right, it is being backported to windows 7 as well. Horrible and >>>> disgusting. >>>> >>>> I already have windows server in LAN dedicated as a update server (work of >>>> my >>>> windows colleagues), so the PC don't have to access windows update servers >>>> outside LAN - this should simplify things. >>>> >>>> Also the PCs must have internet access to email, http, https, ftp, sftp - >>>> simply >>>> the 'usual' stuff. >>>> I think, yet, there should be a way. I'll try to consult mikrotik experts >>>> (the >>>> router brand we use) and guys from our ISP. >>>> If i have something, i'll let you know :-) >>>> >>>> thank you, bb >>>> >>>> Karel >>>> >>>> On 03/05/2016 12:40 AM, Steven Haigh wrote: >>>>> On 05/03/16 07:24, Karel Lang AFD wrote: >>>>>> Hi all, >>>>>> >>>>>> guys, i think everyone heard already about how windows 10 badly treat >>>>>> its users privacy. >>>>> >>>>> My solution to this was to finally rid Windows 7 off my desktop PC - as >>>>> most of the telemetry has also been 'back ported' to Windows 7 also. You >>>>> can't stop it. >>>>> >>>>>> I'm now thinking about a way howto stop a windows 10 sending these data >>>>>> mining results to a microsoft telemetry servers and filter it on our SL >>>>>> 6 linux gateway. >>>>> >>>>> Nope. There are no specific servers in use - just general - so whatever >>>>> you block will end up killing other services. >>>>> >>>>>> I think it could be (maybe?) done via DPI (deep packet inspection). I >>>>>> similarly filter torrent streams on our gateway - i patched standard SL >>>>>> 6 kernel with 'xtables' (iptables enhancement) and it is working >>>>>> extremely well. >>>>> >>>>> I would be interested to see if you could identify telemetry packets in >>>>> the flow - but I'm not predicting much success. If you do get it, make >>>>> sure you let the world know though! >>>>> >>>>>> I read (not sure if true) that some DNS resolutions to M$ servers are >>>>>> even 'hardwired' via some .dll library, so it makes it even harder. >>>>> >>>>> Correct. >>>>> >>>>>> I'm no windows expert, but i'm and unix administrator concerned about >>>>>> privacy of windows desktop/laptop users sitting inside my LAN. >>>>>> >>>>>> What i'd like to come up is some more general iptables rules, than >>>>>> blocking specific IP addresses or names, because, apparently they may >>>>>> change in any incoming windows update ... >>>>>> >>>>>> Anyone gave this thought already? Anyone else's concerned the way i am? >>>>> >>>>> Yup - and as I said, I'm now running Fedora 23 on my desktop (EL lags on >>>>> a few things that I like - so Fedora is a happy medium for me - as I >>>>> still have the fedora-updates-testing repo enabled. My work laptop as >>>>> well as my personal laptop - and now my home desktop all run Fedora 23 >>>>> (KDE Spin if you hate Gnome 3 - like me). >>>>> >>>> >>> >>
Re: samba and ntfs flash drives ???
On 04/03/16 11:05, ToddAndMargo wrote: [...snip...] > # grep denied /var/log/audit/audit.log > type=AVC msg=audit(1457071461.014:2015): avc: denied { write } for pid=26451 > comm="smbd" name="test" dev="dm-1" ino=593703 > scontext=system_u:system_r:smbd_t:s0 tcontext=unconfined_u:object_r:mnt_t:s0 > tclass=dir > > These stem from when I was trying to get SeLinux to work > on the share. "Test" was a shared directory. "Test" > has since been removed. > > I can browse/use the mount point without issue as > long as I do not have an NTFS Flash Drive mounted to it. > > No mention of /mnt/iso in the above > # grep denied /var/log/audit/audit.log | grep iso > # You skipped the 'audit2allow' tip I gave you. - cat | audit2allow type=AVC msg=audit(1457071461.014:2015): avc: denied { write } for pid=26451 comm="smbd" name="test" dev="dm-1" ino=593703 scontext=system_u:system_r:smbd_t:s0 tcontext=unconfined_u:object_r:mnt_t:s0 tclass=dir #= smbd_t == # This avc can be allowed using the boolean 'samba_export_all_rw' allow smbd_t mnt_t:dir write; ----- See the line " This avc can" ... So just do: # setsebool -P samba_export_all_rw 1 -- kind regards, David Sommerseth
Re: samba and ntfs flash drives ???
On 4 March 2016 09:45:38 CET, ToddAndMargo <toddandma...@zoho.com> wrote: >Hi All, > >Google is killing me here! > >Scientific Linux 7.2, 64 bit > >$ rpm -qa samba >samba-4.2.3-11.el7_2.x86_64 > >Is there some trick to mounting an NTFS USB flash drive and >sharing it with Samba? > >I am trying to share an NTFS flash drive with samba. >If the drive is not mounted, I can do what I want >from Windows 7 and XP on the mount point. I have >full access. > >But, when I mount the stick to the mount point and >try to browse the mount with W7 or XP, I get "permission >denied". Specifically, from the W7 machines samba log: > > ../source3/smbd/uid.c:384(change_to_user) > Skipping user change - already user > > ../source3/smbd/open.c:881(open_file) > Error opening file . (NT_STATUS_ACCESS_DENIED) > (local_flags=0) (flags=0) > >I mount suchlike: > ># mount -t ntfs -rw -o >users,exec,sync,uid=todd,gid=users,fmask=000,dmask=000 /dev/sdc1 >/mnt/iso > >(I know I don't need the masks, but I left them there in case >they were needed) > >After mounting: ># ls -al /mnt/iso > >total 1193 >drwxrwxrwx. 1 todd users 4096 Mar 3 23:30 . >drwxr-xr-x. 13 todd users 4096 Mar 3 21:47 .. >-rwxrwxrwx. 1 todd users122 Apr 12 2011 autorun.inf >drwxrwxrwx. 1 todd users 4096 Apr 12 2011 boot >-rwxrwxrwx. 1 todd users 383786 Apr 12 2011 bootmgr >-rwxrwxrwx. 1 todd users 669568 Apr 12 2011 bootmgr.efi >drwxrwxrwx. 1 todd users 0 Apr 12 2011 efi >-rwxrwxrwx. 1 todd users 106768 Apr 12 2011 setup.exe >drwxrwxrwx. 1 todd users 40960 Apr 12 2011 sources >drwxrwxrwx. 1 todd users 0 Apr 12 2011 support >drwxrwxrwx. 1 todd users 0 Apr 12 2011 upgrade > >My smb.conf: > >[iso] > comment = mnt on rn1 -- Mount as M: > path = /mnt/iso > valid users = @users > write list = @users > force group = users > force user = todd > oplocks = no > level2 oplocks = no > strict locking = no > blocking locks = no > force create mode = > create mode = 0777 > force directory mode = > directory mode = 0777 > map system = yes > map hidden = yes > writable = yes > >Trying simpler: > [iso] > comment = mnt on rn1 -- Mount as M: > path = /mnt/iso > force group = users > force user = todd >Doesn't work either > >What am I doing wrong? > >Many thanks, >-T # grep denied /var/log/audit/audit.log If you see something which looks related, pipe them to audit2allow and see what it suggests. Ofen you may get som hints that you need to flip a selinux boolean. -- kind regards, David Sommerseth
Re: "Not using downloaded repomd.xml because it is older than what we have:"
On 13/02/16 19:24, Peter Boy wrote: > Hi all, > > since several months I get constantly from anacreon: >> ——< > /etc/cron.daily/0yum-daily.cron: > > Not using downloaded repomd.xml because it is older than what we have: > Current : Thu Feb 4 16:13:26 2016 > Downloaded: Thu Feb 4 16:13:25 2016 >> ——< > > > The time difference is quite minimal. And a manual „yum update“ > confirms that no updates are waiting. > > Using my favourite search engine I found it might have be caused by > an unresponsive of lazy mirror. But I use the standard configuration, i.e. > the mirror list just includes the three scientificlinux servers. Other > entries refer old bugs long fixed. > > I tried a yum clean all but it didn’t fix it. > > And all our other don’t show this issue, but the configuration is > all the same, at least according to my knowledge. > > > Obviously, there is no harm done and it can be safely ignored. But > it always pulls our issue alert button. Hi, I am seeing exactly the same. I thought it was NTP issues related to my own setup, where I have a local rsync mirror. But then I installed from scratch SL7 on another site without any local mirrors, and the same issue appears there too. So I see this both with public repositories as well as local rsync repositories. I have also seen this on SL6, but not as frequent as on SL7. Even though they cause no obvious harm, it gets quite annoying when you receive many of them during a day ... sometimes even several days in a row. Perhaps yum should be more graceful to the timestamp? And just quiet these messages if the time difference is less than 30 seconds or so. -- kind regards, David Sommerseth -- kind regards, David Sommerseth
Re: two mysteries
On 27/01/16 17:58, Yasha Karant wrote: > My laptop has an external "hardware" expansion insert slot, and I might be > able to find such a 802.11 NIC. USB wireless interfaces should also work, using USB redirection. But it depends on your performance needs on the wireless network. I've done that a few times with USB Ethernet devices when playing around on odd projects. The external "hardware" expansion may or may not work, depending on if is possible to use a PCI Pass-through mode on that interface or not. -- kind regards, David Sommerseth
Re: two mysteries
On 27/01/16 11:13, jdow wrote: >> > Fascinating. I made a bad "assumption" about network devices. It seems they > are created dynamically without any presence in /dev. IIRC, *BSD provides /dev nodes for network devices which the user-space can use for configuring it and such. But it's many years since I played with FreeBSD, so my memory is scarce. -- kind regards, David Sommerseth
Re: two mysteries
On 26/01/16 08:13, Yasha Karant wrote: > On 01/25/2016 04:30 PM, David Sommerseth wrote: [...snip...] >> But KVM is the core hypervior. It is in fact just a kernel >> module which >> you can load at any time on systems with CPUs supporting hardware >> virtualization (VT-d or similar, most modern Intel, AMD and IBM Power 7/8 >> supports KVM). >> >> libvirt is the management backend, which provides a generic API. >> libvirt can >> be used against other hypervisors as well, such as Xen, but probably more >> often used with KVM. >> >> qemu-kvm is the KVM virtual machine process. Each qemu-kvm process is >> started >> per VM. You seldom start these processes manually, but they are >> kicked off by >> libvirt. >> >> virt-manager is a management GUI front-end. And virsh is a console based >> management tool. Both connects to the libvirt API. >> >> Further, you can also download an oVirt Live image and boot that on a >> bare-metal or virtual machine. oVirt can then connect to libvirt and >> provide >> an even more feature rich management tool. >> >> virt-manager and oVirt can also connect to several systems running >> libvirt >> simultaneously, so you can manage more hypervisors from a single >> front-end. >> And there are probably even more front-ends, like "Boxes" (not really >> tried it). >> >> I dunno much about vmware stuff, so I will refrain to comment that. But >> VirtualBox is also two-fold. My experience with VirtualBox is now >> quite old >> (5-6 years ago). You can start VirtualBox guests without a kernel >> support >> module loaded, which would work on most hardware. But performance was >> not too >> good at all. If you got the init.d script to build the kernel module, >> you >> could get quite acceptable performance. However, I see VirtualBox >> more like a >> single package which gives you both hypervisor and management tool in >> a single >> software package. >> >> Even though VirtualBox is more a "single unit" and KVM/Qemu/libvirt >> consists >> of more components ... you normally don't notice that when you start >> VMs via >> the management tools. >> > Thank you for your detailed exposition. My primary concern is that I do > *NOT* want a hypervisor actually controlling the physical hardware; we > have enough security vulnerabilities with a "hardened" supervisor such > as EL 7. You can run virtual machines without a hypervisor. But, that will not give you a good performance in general. Running in this mode is often called 'emulation'. So the hardware a computer needs, is emulated by software in user space, without anything running in kernel space at all. You can do this also with libvirt and qemu too, but then you use 'qemu' and not 'qemu-kvm'. As a related side-track. Running with a hypervisor can only allow guests to be of the same CPU family as the bare-metal host. With emulation, the CPU seen on the inside of the guest can be whatever the emulator supports. With emulation you can run powerpc, mips or even s/390 based environments - but it is slow compared to bare-metal performance - as everything you do is emulated. Likewise with VirtualBox, it goes into emulated mode when it does not have the kernel module (vbox.ko? don't recall right now). This also provides a much poorer performance. I do not know enough about vmware, but their early products did run on hardware before hardware had any virtualization features at all. But I suspect they also needed some kind of kernel module to provide a decent performance. Once the bare-metal hardware got virtualization support, you still need the kernel module - but now the module takes advantage of the hardware capabilities in addition, increasing the performance even more. So to simplify it a bit: Qemu, VirtualBox and vmware (I suspect) needs a kernel module to provide decent performance, and these modules instruments the kernel with at least hypervisor-like capabilities. > My secondary issue is the actual human clock execution time in > the VM as contrasted with the same OS/environment running on the > physical hardware. I have found that current production releases of > VirtualBox and VMware (e.g., VMware player) provide acceptable > performance, although the USB interface on VMware now does seem better > than VirtualBox that evidently still has issues (one of the mysteries). And this is what the hypervisor does. It provides a channel from the hardware on the bare-metal to the guest VM. And to get an acceptable human clock execution time inside a virtual guest OS, you will need a hypervisor. So you have most likely been running both
Re: two mysteries
On 25/01/16 19:32, Yasha Karant wrote: > On 01/24/2016 06:06 PM, Lamar Owen wrote: >> On 01/23/2016 01:30 PM, Yasha Karant wrote: >>> Perhaps someone else has experienced what I related below and can comment >>> -- SL 7x. >>> >>> 1. ... For 802.3, I prefer to use a manual configuration, not >>> NetworkManager. >> >> For a dynamic connection even with a wired Ethernet you should use the >> supported NetworkManager stack, your personal preferences aside. NM works >> and doesn't require munging for a simple DHCP wired connection. >> >>> >>> 2. ...Note that I must use MS Win to work with these devices as the >>> application software for the device in question is *NOT* available for >>> linux, the device is proprietary (no source code available), and >>> CrossOver/Wine does not support USB -- forcing the use of a VM running a MS >>> Win gues >> >> Neither VMware nor VirtualBox ship as part of SL. KVM does, and USB >> passthrough works very well with Windows 7 running in a KVM virtual machine >> on my laptop. It just works, and it's already part of SL; why not use it? >> Performance is very good in my experience, and I'm running a few pieces of >> software in Win 7 for the same reasons as you. You're also far more likely >> to get useful help using KVM, either from the list or from other sources, >> such as the Red Hat or Fedora documentation. > > From the KVM site (http://www.linux-kvm.org/page/Management_Tools) that has a > RedHat logo, there is a list of management interfaces, including VMM (Virtual > Machine Manager -- https://virt-manager.org/screenshots/ ) that also appears > to be a Red Hat entity. Anyone using VMM? VMM appears to allow a true host > OS (supervisor, not hypervisor) with the VM ("hypervisor") running under the > OS (as with VMWare workstation/player or VirtualBox), thus booting an OS, not > a hypervisor that actually provisions for guest supervisors. Is this correct? This was a bit confusing for me (getting late, so probably stupid to reply now). But KVM is the core hypervior. It is in fact just a kernel module which you can load at any time on systems with CPUs supporting hardware virtualization (VT-d or similar, most modern Intel, AMD and IBM Power 7/8 supports KVM). libvirt is the management backend, which provides a generic API. libvirt can be used against other hypervisors as well, such as Xen, but probably more often used with KVM. qemu-kvm is the KVM virtual machine process. Each qemu-kvm process is started per VM. You seldom start these processes manually, but they are kicked off by libvirt. virt-manager is a management GUI front-end. And virsh is a console based management tool. Both connects to the libvirt API. Further, you can also download an oVirt Live image and boot that on a bare-metal or virtual machine. oVirt can then connect to libvirt and provide an even more feature rich management tool. virt-manager and oVirt can also connect to several systems running libvirt simultaneously, so you can manage more hypervisors from a single front-end. And there are probably even more front-ends, like "Boxes" (not really tried it). I dunno much about vmware stuff, so I will refrain to comment that. But VirtualBox is also two-fold. My experience with VirtualBox is now quite old (5-6 years ago). You can start VirtualBox guests without a kernel support module loaded, which would work on most hardware. But performance was not too good at all. If you got the init.d script to build the kernel module, you could get quite acceptable performance. However, I see VirtualBox more like a single package which gives you both hypervisor and management tool in a single software package. Even though VirtualBox is more a "single unit" and KVM/Qemu/libvirt consists of more components ... you normally don't notice that when you start VMs via the management tools. I hope this gave a broader perspective. -- kind regards, David Sommerseth
Re: forced fsck
On 17 January 2016 04:18:22 CET, ToddAndMargo <toddandma...@zoho.com> wrote: >On 01/16/2016 07:10 PM, Brandon Vincent wrote: >> On Sat, Jan 16, 2016 at 8:06 PM, ToddAndMargo <toddandma...@zoho.com> >wrote: >>> Would this be a one time fsck or every boot? >> >> As long as the kernel parameter(s) is passed during boot (manually >> entered or set as the default in GRUB) a fsck should run every time. >> >> Brandon Vincent >> > >Any way to do it once? When you boot the box and see the grub menu, select the proper line with arrow keys and hit the E key on the keyboard; that enters the edit mode. Use arrow keys again to locate the kernel command line and append these options at the end of the line. IIRC correctly, it is ctrl-X to boot with these changes. But these changes are only temporary and only valid for the current boot. -- kind regards, David Sommerseth
Re: Perl 6?
On 10/01/16 23:29, ToddAndMargo wrote: > On 01/10/2016 09:12 AM, James Rogers wrote: >> The usual question: is anyone actually using P6 for anything productive? >> >> --W > > Isn't it a little early for that question considering it > was just released? So the benefit of seeing Perl6 in EL distributions "now" is? Yeah, I know, it can be seen as the "chicken and egg" problem. But that's where EPEL may be the a middle ground, to prove it is needed. However, not everything in Fedora ends up in RHEL either. So don't have too high expectations just yet. There are no reasons RH will spend much time and energy on this before Perl6 gets a critical user mass using Perl6 software RH customers want to use. Just my 2 cents -- kind regards, David Sommerseth
Re: need rsync exclude help
From: ToddAndMargo toddandma...@zoho.com To: SCIENTIFIC-LINUX-USERS SCIENTIFIC-LINUX-USERS@listserv.fnal.gov Sent: 7. March 2015 05:40:43 Subject: Re: need rsync exclude help --exclude='{wine-*,wine-1.7.24}' /home/CDs/Linux /mnt/MyCDs/. I am not real certain that the {} thingy works correctly. Anyway, I only needed 'wine-*' That seems redundant in this case. You can always test such expansions using 'echo' $ echo {wine-*,wine-1.7.24} wine-* wine-1.7.24 $ echo wine-{1.7.24,package-1.2.3} wine-1.7.24 wine-package-1.2.3 Here I also added a little demonstration of how the {} expansion can work. -- kind regards, David SOmmerseth
Re: ypbind not registering with rpcbind on SL7
- Original Message - From: Konstantin Olchanski olcha...@triumf.ca To: Stephen John Smoogen smo...@gmail.com Cc: Dirk Hoffmann hoffm...@cppm.in2p3.fr, SCIENTIFIC-LINUX-USERS SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV Sent: Tuesday, 6 January, 2015 22:18:54 Subject: Re: ypbind not registering with rpcbind on SL7 On Tue, Jan 06, 2015 at 11:20:30AM -0700, Stephen John Smoogen wrote: On 6 January 2015 at 05:08, Dirk Hoffmann hoffm...@cppm.in2p3.fr wrote: I installed SL7 yesterday from the standard DVD in Computing node flavour. yum update ran correctly, then I needed YP/NIS. Wow.. I didn't know ypbind was still in use :)? There is no replacement to NIS for small clusters. Vendors send us in the direction of LDAP, which is supposed to be light weight. Well, if LDAP is light-weight, I hate to see what they consider as normal-weight. With NIS, management is vi /etc/auto.home; make -C /var/yp. Wake me up when LDAP gets anywhere near that easy to use. I'll admit that my IT career has mostly missed the yp/nis days (mostly due to working in companies with just a few handful servers or less). But! I dare you to try out FreeIPA. I've tested it in a slightly bigger environment (~30 boxes), and decided to roll it out at home just for fun to play more with it. It doesn't eat that much CPU or disk resources (well some 100MB), but it is really easy to set up and play with. And with both a reasonable webUI and a command line interface for the same tasks. Firewall and SELinux friendly, and lets you do really nice stuff such as DNS SSHFP (no more need for hosts in ~/.ssh/known_hosts), centralised SSH public key management, Kerberos SSO and all the other stuff NIS can do. Regarding resource usage, at home I installed FreeIPA on an slightly well loaded HP Microserver G7 (AMD N36L) with 8GB RAM running 5 VMs. The average CPU load is 60% and using ~7GB for VMs. And the admin web console works very well and all IPA domain members gets the authentication done fairly quickly. I've not noticed any performance drop on the VMs either. What I basically did: * IPA server - yum install ipa-server - ipa-server-install (see --help for enabling DNS server and more features) - Go to http://$SERVER - Login as admin and start playing * IPA clients to become domain members - yum install ipa-client - Ensure /etc/resolv.conf 'nameserver' points at the IPA server - ipa-client-install (see --help for more advanced features) Also check out the documentation (you'll find relevant versions of it in https://access.redhat.com under Identity Management). It is quite good and accurate. And that's basically it ... run kinit and you have SSO to all your boxes. Or upload your SSH public key to your IPA user account, and you can SSH to all boxes without uploading any public keys anywhere else. My playing has been done with SL6, SL7 and Fedora 19. My next step is to start playing with IPA servers on SL7, which is an even newer version of FreeIPA with some more features. By the way, setting up master-master replication with more IPA servers is also really easy. However, there is a bug in the LDAP server which needs a configuration workaround. But once that's done, it works really smooth. Yes, IPA is probably using more resources than yp/nis, but it also provides much more than just yp/nis. -- kind regards, David Sommerseth
Re: How to do WPA wifi authentication at run level 3 on SL 5.5 ?
On 08/10/14 04:30, Allen Wilkinson wrote: David, Key question is how do I configure network connections with NetworkManager from the command line? Ahh! I see. IIRC, nmcli in EL5 does not support that. I believe there is some support for it in EL6 and even more in EL7. But for EL5, I believe you need to dig into the configuration files in /etc/NetworkManager. I'm sorry, I have only two production boxes left with EL5, and neither of them use NetworkManager, so it's hard for me to point you further. -- kind regards, David Sommerseth On Wed, 8 Oct 2014, David Sommerseth wrote: On 07/10/14 13:55, Allen Wilkinson wrote: I could use help on the SUBJECT problem. This is for an old laptop that uses the ipw2200 wifi driver. It assigns the wifi to eth1. eth0 for wired Ethernet is active okay, and I want eth1 active at the same time. ifup eth1 seems to only allow WEP keys successfully. NetworkManager never seems to connect at run level 3 using WPA for any configuration that I can figure out. nm-tool does show WPA should be possible. Hi Allen, If you already have configured the network connections using NetworkManager, it should be fairly possible to start the wireless network using 'nmcli'. That's a command line tool for Network Manager. You most likely need to play around with 'nmcli con'. F.ex. I have a wirelss config called 'home'. So to connect from the command line, I do this: [user@host:~] $ nmcli con up id home I'll admit, it's a long time since I played with EL5, so it might not be fully supported. But on EL6 and newer, this is possible. -- kind regards, David Sommerseth
Re: How to do WPA wifi authentication at run level 3 on SL 5.5 ?
On 07/10/14 13:55, Allen Wilkinson wrote: I could use help on the SUBJECT problem. This is for an old laptop that uses the ipw2200 wifi driver. It assigns the wifi to eth1. eth0 for wired Ethernet is active okay, and I want eth1 active at the same time. ifup eth1 seems to only allow WEP keys successfully. NetworkManager never seems to connect at run level 3 using WPA for any configuration that I can figure out. nm-tool does show WPA should be possible. Hi Allen, If you already have configured the network connections using NetworkManager, it should be fairly possible to start the wireless network using 'nmcli'. That's a command line tool for Network Manager. You most likely need to play around with 'nmcli con'. F.ex. I have a wirelss config called 'home'. So to connect from the command line, I do this: [user@host:~] $ nmcli con up id home I'll admit, it's a long time since I played with EL5, so it might not be fully supported. But on EL6 and newer, this is possible. -- kind regards, David Sommerseth
Re: [SCIENTIFIC-LINUX-USERS] Questions about SL 7.0
On 03/09/14 10:33, Andreas Mock wrote: Hi Pat, hi Patrick, thanks for your answers and comments. How would someone like me get a SRPM for a binary package found or installed on a SL 7.0 system? yumdownloader --source $PKGNAME -- kind regards, David Sommerseth Von: Patrick J. LoPresti [mailto:lopre...@gmail.com] Gesendet: Dienstag, 2. September 2014 23:22 An: Pat Riehecky Cc: Andreas Mock; SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV Betreff: Re: AW: [SCIENTIFIC-LINUX-USERS] Questions about SL 7.0 On Tue, Sep 2, 2014 at 2:11 PM, Pat Riehecky riehe...@fnal.gov wrote: The sources were taken from git. They were then compared to the sources from the public Release Candidate provided by upstream on April 22 2014. There were very few changes from this Release Candidate to the official release. Nice work. All the Security/Enhancement/Bugfix code comes out of git as the source rpms for these were never publicly released. Does this mean there is no way to correlate security/bugfix updates from Red Hat with the changes in git, and therefore no way to know how far SL is diverging from RHEL over time? Is the git tree entirely RHEL + released updates, or are unreleased CentOS changes mixed in as well? Presumably, anyone with a RHEL subscription (and the right tools) could compare the git repository against the update SRPMs, at least to tell you whether they are the same. Would that be a violation of the subscription terms, I wonder? Just curious. - Pat -- kind regards, David Sommerseth
Re: nmap to find mac addressees
On 02/09/14 20:16, Earl Ramirez wrote: On Tue, 2014-09-02 at 09:47 -0400, Lamar Owen wrote: On 09/01/2014 05:37 PM, ToddAndMargo wrote: An RPM [of autoscan-network] yet to try? In my own quick build attempts, I was not successful in building a running binary from source. Also, the binary tarball didn't run, either. But I'm running 64-bit and the tarball is 32-bit, and there are a few indications that it might not build properly as 64-bit. But it wasn't high on my priority list, either, so I only put half an hour or so into seeing just how easy or difficult it would be to build. I did track down dependencies as they were declared, and the binary built, but it would not actually run as built. So I've put it on the back burner here, since, as I say, it's not a terribly high priority to me at the moment. If you need it, you can use a BackTrack/Kali live disk which as far as I know still includes autoscan-network. I have stumble upon the source RPM for it and I am also in the process of getting it to work on el6 (for now, will be another story for el7). I thinking about moving the install directory from /usr/share/apps to /opt/* directory, thoughts are welcome. I am doing the build on a vanilla el6 install and below are the build dependencies, which I plan to sort out within a few days if time permits. rpmbuild -ba rpmbuild/SPECS/autoscan.spec error: Failed build dependencies: gnomeui2-devel is needed by autoscan-1.50-1.el6.x86_64 libao-devel is needed by autoscan-1.50-1.el6.x86_64 libvorbis-devel is needed by autoscan-1.50-1.el6.x86_64 net-snmp-devel is needed by autoscan-1.50-1.el6.x86_64 gtk+2-devel is needed by autoscan-1.50-1.el6.x86_64 libgtk-vnc-devel is needed by autoscan-1.50-1.el6.x86_64 libgnome-keyring-devel is needed by autoscan-1.50-1.el6.x86_64 vte-devel is needed by autoscan-1.50-1.el6.x86_64 pcap-devel is needed by autoscan-1.50-1.el6.x86_64 This is my hobby and if there anyone with extensive experience I am happy to get some little tips. I'd recommend you to install mock and build it via mock. Your user account must be member of the mock group to function. Mock builds pulls down the needed packages for the distribution you build your packages for, unpacks them in a chroot and does the complete build process inside that chroot. On success, you'll get RPMs and on failures you'll get a lot of log files to study :) As I have my own mock configurations, I don't remember the exact syntax ... but it's something like this: $ mock -r epel-6-x86_64 --rebuild autoscan-1.50-1.el6.src.rpm Sit down and wait for a while, and the results can be found in the /var/lib/mock/epel-6-x86_64/results directory. Using this approach, the same computer can build packages for a vast majority of Fedora and EPEL packages, on several CPU architectures. Look in /etc/mock to see all the available configurations out-of-the-box. Adopting for specific SL builds isn't that hard to accomplish either. Mock is one of the building-blocks koji (the Fedora build system) uses. -- kind regards, David Sommerseth
Re: having trouble downloading with wget
On 21/08/14 19:57, ToddAndMargo wrote: Hi All, Any idea why I can download http://www.uvnc.eu/download/1201/UltraVNC_1_2_0_X64_Setup.exe with Firefox, but not wget? I haven't checked the URLs, but my first hunch is that they do User-Agent filtering on the server side. Try using wget with -U and a user agent string from f.ex. Firefox. -- kind regards, David Sommerseth
Re: wget: need help extracting name
On 29/07/14 19:04, ToddAndMargo wrote: Hi All, I am trying to extract the file name (for the revision) of a file with --content-disposition. I don't actually want the file. Is there a way to have wget tell me the name of the file without actually downloading it? Many thanks, -T This is what I have so far. Tell me the file exists but not its name. wget --spider --content-disposition http://www.overlooksoft.com/packages/download?plat=lx64ext=rpm; Spider mode enabled. Check if remote file exists. --2014-07-29 10:00:38-- http://www.overlooksoft.com/packages/download?plat=lx64ext=rpm Resolving www.overlooksoft.com... 96.127.149.74 Connecting to www.overlooksoft.com|96.127.149.74|:80... connected. HTTP request sent, awaiting response... 200 OK Length: 6737765 (6.4M) [application/x-rpm] Remote file exists. $ curl -I 'http://www.overlooksoft.com/packages/download?plat=lx64ext=rpm' \ | awk -F= '/Content-Disposition: attachment;/ {print $2}' Not sure if wget got a similar feature. I just happen to like curl slightly better. -- kind regards, David Sommerseth
Re: wget: need help extracting name
On 29/07/14 20:37, ToddAndMargo wrote: On 07/29/2014 10:40 AM, David Sommerseth wrote: On 29/07/14 19:04, ToddAndMargo wrote: Hi All, I am trying to extract the file name (for the revision) of a file with --content-disposition. I don't actually want the file. Is there a way to have wget tell me the name of the file without actually downloading it? Many thanks, -T This is what I have so far. Tell me the file exists but not its name. wget --spider --content-disposition http://www.overlooksoft.com/packages/download?plat=lx64ext=rpm; Spider mode enabled. Check if remote file exists. --2014-07-29 10:00:38-- http://www.overlooksoft.com/packages/download?plat=lx64ext=rpm Resolving www.overlooksoft.com... 96.127.149.74 Connecting to www.overlooksoft.com|96.127.149.74|:80... connected. HTTP request sent, awaiting response... 200 OK Length: 6737765 (6.4M) [application/x-rpm] Remote file exists. $ curl -I 'http://www.overlooksoft.com/packages/download?plat=lx64ext=rpm' \ | awk -F= '/Content-Disposition: attachment;/ {print $2}' Not sure if wget got a similar feature. I just happen to like curl slightly better. -- kind regards, David Sommerseth Hi David, I must be missing something. Don't see the name anywhere. And, can't figure out why I get something different with the AWK pipe. -T $ curl -I 'http://www.overlooksoft.com/packages/download?plat=lx64ext=rpm' HTTP/1.1 403 Forbidden Date: Tue, 29 Jul 2014 18:34:31 GMT Server: Apache Content-Type: text/html; charset=iso-8859-1 $ curl -I 'http://www.overlooksoft.com/packages/download?plat=lx64ext=rpm'| awk -F= '/Content-Disposition: attachment;/ {print $2}' % Total% Received % Xferd Average Speed TimeTime Time Current Dload Upload Total SpentLeft Speed 0 00 00 0 0 0 --:--:-- --:--:-- --:--:-- 0 Odd! You get 403 Forbidden. I don't get that from my Fedora 19 box: $ curl -I 'http://www.overlooksoft.com/packages/download?plat=lx64ext=rpm' HTTP/1.1 200 OK Date: Tue, 29 Jul 2014 20:14:31 GMT Server: Apache X-Powered-By: PHP/5.4.28 Pragma: public Expires: 0 Cache-Control: public Content-Description: File Transfer Content-Disposition: attachment; filename=overlook-fing-2.2.rpm Content-Transfer-Encoding: binary Content-Length: 6737765 Vary: User-Agent Content-Type: application/x-rpm However on a SL6.5 box, I do get 403 Forbidden. I tried changing the User-Agent string, and that helped. Seems some curl versions have been banned on that site. Pick a user-agent string from here: http://www.useragentstring.com/pages/Firefox/ And run curl with '-A $USERAGENT' ... that should fix it. -- kind regards, David Sommerseth
Re: SL REST API
On 27/05/14 18:52, Yasha Karant wrote: Does SL (i.e., TUV EL) have a standard enterprise-quality production REST API that will interoperate with non-EL clouds? The most I could find on a short search is: http://developerblog.redhat.com/2013/12/12/advanced_integration_rhevm-part1/ Advanced integration with Red Hat Enterprise Virtualization Manager (RHEV-M) – Part 1 of 2 and https://fedorahosted.org/rhevm-api/ This is an effort to define an official REST API for Red Hat Enterprise Virtualization http://www.redhat.com/virtualization/rhev/. but that the fedorahosted project above is obsolete, replaced by: http://www.ovirt.org/Subprojects in which any mention of TUV by name is in the title of each reference. Hi, I think you might find deltacloud interesting. http://deltacloud.apache.org/ -- kind regards, David Sommerseth
Re: ssh -X xinit failure
On 07/05/14 04:33, Yasha Karant wrote: Thanks for the information. At my institution, we were told by the university network security group that after ssh -X, one still needed to activate X for the session by xinit or the like for security reasons. Evidently, the persons were thinking of some other environment (MS Windows perhaps?). Indeed, xeyes and firefox both work fine from the remote host to the local client workstation. A question: as a regular X window manager desktop from the remote machine is not displayed (that is, the pull down menu Applications under Gnome or the equivalent from KDE), is there any mechanism to get such a menu, etc., displayed? What is the default GUI file manager (that allows an end user to point and click on an executable file to execute the application) that can be invoked from a remote terminal? Running this over ssh will most likely not work well at all. If you want a remote desktop experience, look into nomachine or freenx: https://www.nomachine.com/ http://wiki.centos.org/HowTos/FreeNX Another alternative is to start Xvnc and tunnel the VNC port Xvnc establishes from your remote server via SSH. Then use a local VNC client to connect to the same port. This may work, but may also be worse than nomachine. Using anything else, will most likely just cause grief and frustration. The X11 protocol isn't easily tunnelled, and requires quite some stable bandwidth to work decent. -- kind regards, David Sommerseth
Re: Security ERRATA Important: openssl on SL6.x i386/x86_64
On 09/04/14 16:42, Pat Riehecky wrote: This is a reminder of this security errata. Any SL6 system should apply this update. If your system has been applying security errata regularly it is vulnerable until this update is applied. Systems with yum-autoupdate enabled using the default configuration have the update applied and only need to restart applications linked against openssl. All applications linked against openssl must be restarted for this update to take effect. I installed lsof on my boxes and used [root@host: ~]# lsof | grep -E libcrypto|libssl | grep DEL to identify processes/services which needs to be restarted. -- kind regards, David Sommerseth
Re: Any 7 rumors?
On 09/04/14 16:27, Paul Robert Marino wrote: No it was always required because the shopping cart itself may in some cases contain data which could possibly be used to gain access to sensitive customer data. Also in a sense data about who purchases what and where could also be used to mask credit card fraud by making the fraudulent charges look like the normal shopping activities of the card holder. Really!? I've been involved in a few PCI-DSS certification rounds for a company which provided online payment services back in the days. Granted that's some years ago now (2005 to 2008-ish). Even though our scope was limited to only processing credit card information, we did not see any requirements anywhere at that time for the shopping cart to be PCI-DSS certified. In fact one of our sales arguments at that time was that our customers could avoid certifications by implementing our online payment terminal. We even had some discussions with our auditor about this, who gave his blessings to our product. The solution we provided in this case would take care of retrieving the credit card information from the customer, process the payment and just provide a status back to the merchant. Merchants using a payment API for processing payments would in some cases need certification, based on the amount of transactions they had; this I believe has become much stricter since those days. And just to have mentioned it, the solutions we provided was based upon Gentoo(!) servers. We even got very positive feedback for having absolutely minimum installs on our production servers, plus kudos for our maintenance routines. Of course, many of the requirements have most likely changed since then. But I don't recognise the always required in regards to shopping carts. -- kind regards, David Sommerseth On Wed, Apr 9, 2014 at 8:13 AM, James M. Pulver jmp...@cornell.edu wrote: We were recently informed PCI compliance also extends to the shopping cart software, this may be new this year... -- James Pulver CLASSE Computer Group Cornell University From: owner-scientific-linux-us...@listserv.fnal.gov [mailto:owner-scientific-linux-us...@listserv.fnal.gov] On Behalf Of Paul Robert Marino Sent: Tuesday, April 08, 2014 11:26 PM To: Nico Kadel-Garcia; ToddAndMargo Cc: Scientific Linux Users Subject: Re: Any 7 rumors? Well frankly if you need PCI-DSS compliance pay for RHEL. Its honestly not that expensive for the few systems that really require it. Only the system's that handle credit cards supposedly require it and in most ecommerce companies that's probably 2 to 4 system's so what's the problem wit paying $750 a year each for those few systems to not have to deal with the problems and giving the stock investors a warm and fuzzy feeling. Your time spent on it costs them more money and ti reduces all the stress on every one if you buy compliance on the cheap. -- Sent from my HP Pre3 On Apr 8, 2014 22:55, Nico Kadel-Garcia nka...@gmail.com wrote: On Tue, Apr 8, 2014 at 10:14 PM, ToddAndMargo toddandma...@zoho.com wrote: Hi All, I have a customer who is going to have to upgrade a whole pail of stuff for PCI compliance (credit card security). Part of what he is going to have upgrade is his old CentOS 5.x server (it is too underpowered to handle his new software along with the addition drag caused by adding File Integrity Monitoring [FIM] Software). Any rumors as to when EL 7 will be out? Many thanks, -T Shortly after our favorite upstream vendor publishes it? I don't see the relevance though. If he needs to update CentOS 5, update it to SL 6 or CentOS 6. Why wait for RHE 7 to update? It's going to be major cluster futz with the the switch tu systemd from init scripts, with /bin being migrated to /usr/bin, and the other major changes. It will be much simpler, and much, much safer, to update to CentOS 6 or SL 6 first!
Re: How do I create a link from vfat to ext4?
On 04/03/2014 08:52 PM, Patrick J. LoPresti wrote: You can use a bind mount to make an ext4 directory appear in the VFAT tree. Do man mount and search for bind, and/or search the web for bind mount. - Pat Hi Pat, Thank you! From man page, it sounds like I am writing to the two different locations at the same time: After this call the same contents is accessible in two places Is this just a play on words? Bind mounts are special. It basically mounts an already mounted directory yet another place. Say you have this scheme: /dev/sda4 - /mnt/mydata /dev/sdb2 - /mnt/friendsdata If you add a 'friends' directory in /mnt/mydata ... giving you /mnt/mydata/friends, and the do bind mount: mount -o bind /mnt/friendsdata /mnt/mydata/friends This results in that you have access to the same data in both /mnt/friendsdata and /mnt/mydata/friends ... But all data is read and written from/to /dev/sdb2. It's just that you have loaned an already mounted directory into your /mnt/mydata directory. These bind mounts are kind of a I want what you have-mount. Bind mounts are particularly handy when you work with chroots and wants to grant access to certain files outside the chroot, where symlink is impossible. With bind mounts, you can also the same with files; not just directories. I hope this clarified a bit. -- kind regards, David Sommerseth
Re: Need help with rpm rebuild error
On 25/03/14 03:03, ToddAndMargo wrote: On Mon, Mar 24, 2014 at 9:48 PM, ToddAndMargo toddandma...@zoho.com wrote: Hi All, SL 6.5, 64 bit The following does not rebuild (fc15 or fc20): # rpmbuild --rebuild clipit-1.4.2-5.fc20.src.rpm /root/rpmbuild/BUILDROOT/clipit-1.4.2-5.fc20.x86_64/etc/xdg/autostart/clipit-startup.desktop: error: value GNOME;XFCE;LXDE;Unity;MATE; for key OnlyShowIn in group Desktop Entry contains an unregistered value MATE; values extending the format should start with X- Error on file /root/rpmbuild/BUILDROOT/clipit-1.4.2-5.fc20.x86_64/etc/xdg/autostart/clipit-startup.desktop: Failed to validate the created desktop file Any way around this MATE error? Many thanks, -T On 03/24/2014 06:58 PM, Nico Kadel-Garcia wrote: Fedora has a significantly more recent version of RPM than Scientific Linux 6, and many of the dependencies have significantly modified names. That seems to be a stable package, it was in Fedora 15. Perhaps you could rebuild the SRPM from a Fedora 15 archive site, or compare the .spec files, to look for the changes from the older version? Hi Nico, You missed the fc15 in does not rebuild (fc15 or fc20). First thing I tried. Same exact error. RATS! Bear in mind that EL6 is based upon Fedora 12-13 (roughly). So you'll need to figure out which packaging features have been changed since that time. But I believe this is more related to an issue inside the clipit-startup.desktop file than an RPM .spec file issue - most likely related to xdg-utils (Portland). -- kind regards, David Sommerseth
Re: gstreamer1-plugins-base
On 06/03/14 10:18, Ian A Taylor wrote: Sir/Madam Does anyone know how I can install the package gstreamer1-plugins-base on SL 6.5 So I can playback video mp4 files I can highly recommend the Fluendo plug-ins, if you want to stay on the legal side (which is a requirement for the computers I use for work). Fluendo are heavily involved in gstreamer, and they provide gstreamer modules to playback formats which are patented. It costs a little bit, but the cost isn't that bad. And renewals are even more affordable. http://eu.fluendo.com/shop/product/oneplay-codec-pack/ For my part, to support their open source work on the gstreamer core, I don't mind paying a little bit. -- kind regards, David Sommerseth
Re: running SL6.5 in emulator qemu-system-x86_64
On 05/03/14 18:23, Boryeu Mao wrote: I did use qemu-system-x86_64 on the SL65 image. I will first try qemu-kvm, and then (as a lst resort) virt-manager and libvirtd - I had to do quite a bit work to get libvirt going when I was using Debian/Knoppix, and the relative simplicity of qemu-system-x86_64 (just 'qemu-system-x86_64 -cdrom CDROM.iso') was quite nice. In any event the advantage of libvirt may be worth the work. My experience with virt-manager on Fedora, RHEL and SL the last 3-4 years have been that it's quite capable as a desktop virtualisation tool. The pacakges available out of the box have sane default configs to just get started (tm). If you want to use more advanced features, virsh is the command line interface to libvirt. But having a little idea of the features available in virt-manager can help understanding virsh better. -- kind regards, David Sommerseth On 3/5/14, David Sommerseth sl+us...@lists.topphemmelig.net wrote: On 05/03/14 16:32, Boryeu Mao wrote: I have recently migrated from Debian-based Knoppix to SL (livecd-iso-to-disk from SL-65-x86_64-2014-02-06-LiveDVD.iso), and installed the emulator from the package qemu-system-x86-1.0-27.2.x86_64.rpm). The emulator booted a Knoppix image without incidents (albeit somewhat slowly, KNOPPIX_V7.2.0DVD-2013-06-16-EN.iso). However, booting the SL 65 LiveDVD didn't go so well - it hangs in the startup screen with the inner arc stopping at about 3 o'clock position. I wonder if there may be another package for qemu that I should use, or if there is another virtual machine that may be more robust for SL. Thanks in advance for help or suggestions. SL-65-x86_64-2014-02-06-LiveDVD.iso (via livecd-iso-to-disk) HP laptop Pavilion g6 with dual AMD A6-4400M APU AFAICT, AMD A6-4400M supports hardware virtualization (aka KVM in the Linux world). I'd recommend you to try using qemu-kvm instead of qemy-system-x86. That should improve the performance noticeably. In addition, the SL image you've downloaded is 64 bits. I don't know if you try to run that via qemu-system-i386 or qemu-system-x86_64. The former is 32bit /emulation/, the latter 64bit. But only qemu-kvm uses virtualization. Try also to kick off your images using virt-manager. It's usually fairly simple to configure new VMs that way, and just point it to your images. This way, virt-manager/libvirtd will take care of setting up the proper arguments to qemu-kvm, for better performance. Also try to use the virtio drivers wherever possible, as they don't require the complete hardware emulation layer. -- kind regards, David Sommerseth
Re: server side spam filters
On 11/02/14 02:13, Yasha Karant wrote: Our site has been edicted to Microsoft Exchange server with a Barracuda spam filter. There are numerous difficulties, one of which is spam not being filtered and non-spam being so filtered (significant increase in mission critical false positives). At present, the administrative authorities (all of whom appear to be management professionals, not internals nor systems folks) insist on Exchange, allowing open systems standards compliant end-users to have IMAP service. Given this, what are the best server-side spam filters, either hardware or software? Best should be based upon current field-deployed experience and/or unsolicited external reviews (not vendor-supported independent reviews). I've put up a fairly simple Postfix + Amavis-new + SpamAssasin server in front of some of my Zimbra servers to get rid of the worst trash (we also had some other requirements too, but that's not important in this thread). I configured Postfix with several RBLs, SPF and postgrey. In addition I added these smtpd_recipient_restrictions: reject_unknown_reverse_client_hostname, reject_invalid_hostname, reject_non_fqdn_hostname, reject_non_fqdn_sender, reject_non_fqdn_recipient, reject_unknown_sender_domain, The RBLs I have had great success with are: reject_rbl_client bl.spamcop.net, reject_rbl_client zen.spamhaus.org, reject_rbl_client bl.blocklist.de, reject_rbl_client b.barracudacentral.org, reject_rbl_client bl.spamcannibal.org, reject_rbl_client cidr.bl.mcafee.com, The two first ones and barracudacentral.org seems to be those being triggered most. Barracudacentral requires a registration (they want the IP of your DNS resolver doing the queries). With all this in place, I reduced the spam which SpamAssassin filtered out from 75-80% to ~20-25%. I had to remove SORBS, as they actually listed a lot of valid SMTP relays ... and for those companies being hit here, it was just a too costly operation to fix each time it happened. On the other hand, the other RBLs catch quite fine what SORBS blocked correctly. In regards to SPF, that works pretty well. I did it even stricter than the default configuration (I use python-policyd-spf), where I set PermError_reject = True. That enforces that SPF rules which are explicit much harder. And with postgrey, I learned that you need at least a 10 minutes threshold. For one of the servers I maintain, postgrey blocks ~25% of all mail attempts. On antoher one (low traffic), the hit rate was so low I actually removed. So you need to test and see if it can match your needs. -- kind regards, David Sommerseth
Re: Ping Nico: AD?
On 10/02/14 04:19, ToddAndMargo wrote: On 02/09/2014 02:45 PM, Nico Kadel-Garcia wrote: On Sun, Feb 9, 2014 at 2:45 PM, ToddAndMargo toddandma...@zoho.com wrote: I take it old-out-of-date (SL) isn't supporting Samba 4 yet. Nope, it's in Fedora and RHEL 7 beta and places .like my github repo. What would you think of just doing a fedora 20 server, instead of suffering with all the out of date stuff on RHEL? I did that once. As it took quite a long time before CentOS 6 came out (before I got to know SL, but SL6 was also not shipped). It worked, kind of. But would I do it again? Probably not. I installed Fedora 11, and it was regularly tons of updates. At some point I did an upgrade to Fedora 12 (with massive downtime, due to the Fedora upgrade process - at that time, Anaconda could upgrade via the ISO image). But after that, I stayed far too long on Fedora 12. This server was a host for KVM guest hosts. But I also stayed on Fedora 12 on purpose, as EL6 was based on F12-F13. So what I ended up doing was a scratch install of SL6 on some of the free space I had on the LVM (on a completely separate VG). Reused /boot without reformatting it (so I could fallback to F12 if needed). Copied over the needed /etc configs from F12 to SL6, and modified them as needed too (not too much work, but took some time). And then booted SL6 ... And that actually worked out far better than I ever dared hoping for. Having this said. I did this process on my private laptop (which also had been stuck on F12 on purpose too), where I also run KVM guests. So I gained quite some experience and felt comfortable enough to dare this on a production server. I don't regret moving to SL6 at all. So if you have time to wait for SL7, wait for it. If you're in a pinch and must have this server running yesterday, consider this approach, but I would probably stay on Fedora 19 for a little while - but I would probably try to stay cool as long as possible. The good thing with Fedora right now is that they're figuring out the process of Fedora.next. So Fedora 21 will most likely not come at normal schedule (which would be sometime around Q4 probably), but somewhat later. Which means F19 will not be EOL for a little bit longer. -- kind regards, David Sommerseth
Re: Ping Nico: AD?
On 09/02/14 20:45, ToddAndMargo wrote: Question: what do you see as an advantage of Samba's AD over just using Samba as an old fashioned Domain Controller? A side track actually. If you're looking for something like AD for Linux, which can also integrate with AD (in version 3) ... Have a look at the FreeIPA project. I'm also believe FreeIPA integrates with Samba4 too. http://www.freeipa.org/page/About Not sure how it is with RPM packages/yum repos for FreeIPA, though. I take it old-out-of-date (SL) isn't supporting Samba 4 yet. In SL6, there are samba4 packages available. I believe it is a tech-preview in RHEL6. -- kind regards, David Sommerseth
Re: Autoupdate
On 31/01/14 16:43, CS DBA wrote: On 1/31/14, 8:39 AM, Jose Marques wrote: On 31 Jan 2014, at 15:38, David Sommerseth sl+us...@lists.topphemmelig.net wrote: If you have enabled the sl6x repositories, then yes, this is the expected behaviour. Is there a way of having auto-update and retaining the old behaviour? The University of St Andrews is a charity registered in Scotland, No. SC013532. # yum erase yum-autoupdate This might be the solution if you want to disable all auto-updates. But this is a far more drastic move than just disabling the sl6x repositories. Then you will stay always updated on the same SL minor release, until you do the upgrade manually. -- kind regards, David Sommerseth
Re: Finally figured out what is causing bug 836696
On 10. des. 2013 20:43, Jeff Siddall wrote: On 12/09/2013 11:20 PM, ToddAndMargo wrote: then you absolutely want to be running them against a snapshot rather than a live FS and LVM makes this easy. Never really cared for LVM. Always used the direct partition approach. Well, perhaps I can try to convince you some more. Take another example of upgrading to a bigger disk. Huge PITA if you use direct partitions. Shut the system down and use a live OS or something while you move over all the data -- which could take hours or days depending on what you need to move. If you are really obsessive you probably want to make sure nothing got lost in the move so there is a whole compare exercise after it finishes. If you have LVM you simply install your new drive (assuming you can hotswap you don't even have to shutdown for that) run pvcreate, vgextend and then pvmove. Some hours (or days) later it finishes and your data is magically on the new disk without even a moment of downtime. A lvextend and a resize2fs (or whatever utility resizes the FS you use) and you can start using the extra space, still with no downtime. That is pure sysadmin gold! BTW: I also use LVM on my offsite backup disks. I just use the same volume group/volume name on all the disks. Works with LUKS also. I'll even recomend fsadm, which makes resizing live filesystems even easier and safer. I've even done resize of / on a running system without any issues. The fsadm utility will do much of the filesystem tasks for you, in a safe an controlled manner. # fsadm -v -l resize /dev/vg.../lv... $NEWSIZE If you're more cautious, you can add -n (--dry-run) and even do a 'fsadm check' first. This takes care of resizing all the needed pieces. But 'fsadm check' can only be done on an unmounted volume, iirc. I have not tried fsadm on a direct partition, though. It might also work there. These LVM features can also be very useful if you're using virtual machines with LVM, as adding and removing virtual drives on-the-fly is very easy in such environments. kind regards, David Sommerseth
Re: Finally figured out what is causing bug 836696
On 10. des. 2013 21:27, Bluejay Adametz wrote: Never really cared for LVM. Always used the direct partition approach. Well, perhaps I can try to convince you some more. I never used LVM either, but in my defense, these were/are factory systems where it was extremely unlikely that the system would grow and require more storage, and if one did, we'd have to remove the smaller drives and put in bigger ones. This is actually one typical area where LVM would really make life easier with shorter downtime. But it of course depends on the amount of data. Without hotswap, you would of course need a reboot to add the new hard drive and another one later on to remove the old old. But with hotswap capable hardware you could do this job with 0 downtime. kind regards, David Sommerseth
Re: How do I reset my default router without rebooting?
On 05. des. 2013 06:41, James Rogers wrote: Someone said: P.S. the route command is a legacy command from the 2.2 kernel days and should not be used any more. Did they mean the 'route' command or is the 'ip route' hierarchy now depricated? If we're not supposed to use the ip commands what are we supposed to use now? In the Linux world iproute2 (that is, the 'ip' command) seems to be the preferred utility these days. However net-tools (which is ifconfig, route, netstat, arp, etc, etc) is still shipped in basically all Linux distros. And if you also look at FreeBSD, they still cling to net-tools as the primary tool. I believe the same is for NetBSD and OpenBSD too. iproute2 does have quite some advantages on Linux over net-tools, like better IPv6 configuration, secondary IP addresses without using interface aliases, slicker tunnel configurations to mention a few. I personally prefer the configuration syntaxes in iproute2 over net-tools these days. But it's taken me a little while to get used to output of iproute2. But I still wouldn't expect net-tools to disappear any time soon. Net-tools is still shipped in Fedora and other more bleeding edge Linux distros. -- kind regards, David Sommerseth
SELinux de-mystified
Hi all, As there has been a couple of SELinux discussions lately, I thought this article could help explain better what SELinux is all about and why it's such a great tool. It's written by one of the core SELinux guys in Red Hat, Daniel Walsh and illustrations by one of Fedora's UX designers, Máirín Duffy. Your visual how-to guide for SELinux policy enforcement http://opensource.com/business/13/11/selinux-policy-guide Hope you'll enjoy the reading :) -- kind regards, David Sommerseth
Re: SELinux de-mystified
On 13. nov. 2013 20:26, John Musbach wrote: Maybe it's just me, but it seems like a serious failing of SELinux's efforts when most people I've encountered in the Linux world have the policy of just disabling SELinux in their images. Not sure if this was intended as a fire torch or not, or I'm just being a bit sensitive. But I can turn it around: IPv6 has been available for over a decade (if not longer). Is it a failure of IPv6 that so few enables IPv6 in their networked environments?. Of course not. It's about convenience and resistance of changing your attitude to new technologies. But eventually you're forced to take the step. And it's been a similar situation with iptables (and firewalling in Windows, for that matter). People were mostly ignorant to the concept of firewalling, until they realised they had to implement it to have a more secure environment. Is iptables (or firewalling) considered a failure today? During EL6 installation, there is no way you can disable SELinux. It needs to be done explicitly afterwards. This is because SELinux is considered to work so well most users really don't need to think about it. Seriously. SELinux has also been available since EL4 and Fedora Core 3. SELinux is celebrating 10 years these days. It's not something brand new, but it is beginning to really gain traction. These days even SEAndroid is on the way (that is SELinux for Android). SELinux has developed a lot, and is far more easily available and usable today than it was 10 years ago. Please don't be afraid of it! To all of you SELinux sceptic, I have only this to say: If you first grasp the concept of labelling, SELinux isn't much more difficult than what iptables used to be in the beginning. And that article from Dan Walsh gives a very easy to understand introduction to SELinux. And seriously, unless you really have a really odd setup, SELinux will in not give you any troubles in EL6. I have set up roughly 20 different SL6.x servers the last years. I can't remember having had any real issues related to SELinux. This has been everything from LDAP servers, web servers (apache and nginx), e-mail servers (both postfix+amavis+spamassasin and Zimbra), database servers (both PostgreSQL and MySQL). I honestly can't remember having had much troubles with SELinux at all. If SELinux did kick in, it was usually just to flip some SELinux booleans (semanage boolean --list), modifying some network ports context (semanage port --list) or adding some extra paths for correct file labelling (semanage fcontext --list). Changing those things are really not more difficult than adding additional iptables rules. And to figure out if it is SELinux to blame: grep denied /var/log/audit/audit.log Really, stop disabling it! Try it for real and embrace SELinux now! kind regards, David Sommerseth On Nov 13, 2013, at 2:17 PM, David Sommerseth sl+us...@lists.topphemmelig.net wrote: Hi all, As there has been a couple of SELinux discussions lately, I thought this article could help explain better what SELinux is all about and why it's such a great tool. It's written by one of the core SELinux guys in Red Hat, Daniel Walsh and illustrations by one of Fedora's UX designers, Máirín Duffy. Your visual how-to guide for SELinux policy enforcement http://opensource.com/business/13/11/selinux-policy-guide Hope you'll enjoy the reading :) -- kind regards, David Sommerseth
Re: Anyone know of a best ISO VM for security testing?
On 11/09/13 19:03, Todd And Margo Chester wrote: Hi All, I am getting tooled up to do some Penitration Testing for PCI compliance (Ethical Hacking). Refernce: https://www.pcisecuritystandards.org/pdfs/infosupp_11_3_penetration_testing.pdf It's a long time since I've looked at such tools, but I vaguely remember Backtrack-Linux to be quite state of the art. Not sure if it still is. http://www.backtrack-linux.org/ kind regards, David Sommerseth
Re: Robust local mirroring of SL
On 03/09/13 12:47, John Rowe wrote: Until now there seems to have been no _robust_ way of using a local repository for yum updates. Editing the .repo files, as advised by the HowTo, breaks updates of the repo files themselves, and each of the methods on yum's own web page has a list of disadvantages. Using a local squid proxy has problems with mirroring. But SL6x would finally seem to provide a simple way to do this! From man yum.conf: As of 3.2.28, any file in /etc/yum/vars is turned into a variable named after the filename This means that sl6x.repo can contain baseurl entries such as: baseurl=http://ftp.scientificlinux.org/linux/scientific/6x/$basearch/updates/security/ $localrepo/6x/$basearch/updates/security/ http://ftp1.scientificlinux.org/linux/scientific/6x/$basearch/updates/security/ ftp://ftp.scientificlinux.org/linux/scientific/6x/$basearch/updates/security/ and the yum-conf-sl6x rpm could contain the above SL6x.repo plus a file /etc/yum/vars/localrepo consisting of the single line: http://ftp2.scientificlinux.org/linux/scientific To use a local repository we just replace the content of /etc/yum/vars/localrepo with the URL of the local repository, eg: file:///somewhere/SL/scientific The change is simple, persistent across upgrades and doesn't break the repo files. This raises the questions: 1. Am I the only one finding using a local repository to be irksome, or is everybody else just smarter than I am? 2. Would the SL maintainers be willing to put something like this into the yum-conf-sl6x RPM? Hi John I'm having a local mirror of SL, EPEL and a few more. They're basically just rsync mirrors. To enforce the use of these mirrors, I've just added my own RPM package which I've installed on all my systems. This package contains the needed .repo file with pointers to the mirrored repositories. And in one of the RPM spec file scriptlets (forgot which one, too lazy to dig it up) I'm calling yum-config-manager to disable the other repositories if they are available. With this approach, I've not had any issues at all. And my kickstart scripts ensures installing this package as early as possible. I've also added the same repositories with 'repo' statements in the kickstart files, which ensures quite quick install of new systems with all the latest bits. I think one of the advantages here is that the original repository files are not modified (except the disablement). Which means it's quick and easy to rollback in case of some critical issues with the repo server. And if I decide to add or remove some of the mirrors, I just spin a new package with the updated repo files (and yum-config-manager statements) and they gets installed automatically on all systems. kind regards, David Sommerseth
Re: Replacement for Acrobat Professional?
On 11/06/13 06:59, Yasha Karant wrote: The only thing missing is Adobe Distiller that converts a PostScript file to PDF. You have ps2pdf from the ghostscript RPM which can do that for you. Only commandline, though. -- kind regards, David Sommerseth
Re: [SCIENTIFIC-LINUX-USERS] is this a this virus or an error
On 28/05/13 21:36, Pat Riehecky wrote: On 05/28/2013 02:08 PM, Yasha Karant wrote: The latest ClamAV that I can find pre-ported fro SL6 x86-64 is http://pkgs.repoforge.org/clamav/clamd-0.97.7-1.el6.rf.x86_64.rpm EPEL has a slightly newer version of this package: http://koji.fedoraproject.org/koji/buildinfo?buildID=413926 I can recommend the EPEL version. I'm using it in several production environments. Will this RPM override dependencies in the stock SL distribution? EL (and Linux in general) does not seem to have reliable polymorphism -- the default for these sorts of dependencies generally does not seem to install a different executable/library sub-tree independent of the stock distribution except in so far as the same files (e.g., libraries) are used. There should be no issues with running the EPEL version at least. No dependency issues from what I've seen. However, ClamAV still appears to be pre-production (0.x, not 1.x). Is it stable and useful? Don't look yourself blind on version numbers - especially on Open Source project. Having a release being 0.x-something doesn't mean it's not stable or usable. You need to check what the each project defines as their stable release. http://www.clamav.net/lang/en/ And FWIW, Zimbra ships with these ClamAV 0.x versions as well as part of the integrated anti-virus mail checks. What is even more important with anti-virus, is updated virus databases. Which gets updates almost every day, and freshclam in ClamAV takes care of that job. -- kind regards, David Sommerseth
Re: PYCURL ERROR 7 - couldn't connect to host
On 29/05/13 12:38, Semi wrote: Yum doesn't work on SL6.4, it get error and quit. yum update Loaded plugins: refresh-packagekit, security http://dl.atrpms.net/el6-x86_64/atrpms/stable/repodata/repomd.xml: ^ This isn't really related to SL at all, it's the ATrpms repo not being functional. Try getting in touch with them instead. I presume this would be a good place to start: http://lists.atrpms.net/pipermail/atrpms-users/ http://lists.atrpms.net/mailman/listinfo/atrpms-users -- kind regards, David Sommerseth
Re: PYCURL ERROR 7 - couldn't connect to host
On 29/05/13 14:10, Semi wrote: easy_install of python doesn't work too on SL6.4. easy_install python-Levenshtein Searching for python-Levenshtein Reading http://pypi.python.org/simple/python-Levenshtein/ Download error: [Errno 111] Connection refused -- Some packages may not be found! Reading http://pypi.python.org/simple/python-Levenshtein/ Download error: [Errno 111] Connection refused -- Some packages may not be found! Couldn't find index page for 'python-Levenshtein' (maybe misspelled?) Scanning index of all packages (this may take a while) Reading http://pypi.python.org/simple/ Download error: [Errno 111] Connection refused -- Some packages may not be found! No local packages or download links found for python-Levenshtein error: Could not find suitable distribution for Requirement.parse('python-Levenshtein') Come on! Can you please read the error messages more carefully? It is pretty obvious it cannot connect to pypi.python.org in this case. That is also not a Scientific Linux related server. -- kind regards, David Sommerseth
Re: PYCURL ERROR 7 - couldn't connect to host
On 29/05/13 14:16, Semi wrote: You can try links, all of them alive. Something happened with Linux. Please check this problem carefully. -- $ curl -k -D- https://pypi.python.org/simple/python-Levenshtein/ HTTP/1.1 200 OK Date: Wed, 29 May 2013 12:17:28 GMT Server: nginx/1.1.19 Content-Type: text/html; charset=utf-8 Strict-Transport-Security: max-age=31536000 Cache-Control: max-age=3600, public Via: 1.1 varnish Content-Length: 800 Accept-Ranges: bytes Via: 1.1 varnish Age: 969 X-Served-By: cache-s34-SJC2, cache-a16-AMS X-Cache: MISS, HIT X-Cache-Hits: 0, 3 X-Timer: S1369828879.616590023,VS0,VS77,VE129,VE969233 Vary: Accept-Encoding htmlheadtitleLinks for python-Levenshtein/titlemeta name=api-version value=2 //headbodyh1Links for python-Levenshtein/h1a href=../../packages/source/p/python-Levenshtein/python-Levenshtein-0.10.2.tar.gz#md5=c8af7296dc640abdf511614ee677bbb8 rel=internalpython-Levenshtein-0.10.2.tar.gz/abr/ a href=http://trific.ath.cx/resources/python/levenshtein/; rel=homepage0.10.1 home_page/abr/ a href=http://github.com/miohtama/python-Levenshtein; rel=homepage0.10.2 home_page/abr/ a href=http://webandmobile.mfabrik.com;http://webandmobile.mfabrik.com/abr/ a href=http://celljam.net/;http://celljam.net//abr/ a href=http://github.com/miohtama/python-Levenshtein/tree/;http://github.com/miohtama/python-Levenshtein/tree//abr/ /body/html -- This works ... you probably have a local network issue. -- kind regards David Sommerseth On 29-May-13 15:13, David Sommerseth wrote: On 29/05/13 14:10, Semi wrote: easy_install of python doesn't work too on SL6.4. easy_install python-Levenshtein Searching for python-Levenshtein Reading http://pypi.python.org/simple/python-Levenshtein/ Download error: [Errno 111] Connection refused -- Some packages may not be found! Reading http://pypi.python.org/simple/python-Levenshtein/ Download error: [Errno 111] Connection refused -- Some packages may not be found! Couldn't find index page for 'python-Levenshtein' (maybe misspelled?) Scanning index of all packages (this may take a while) Reading http://pypi.python.org/simple/ Download error: [Errno 111] Connection refused -- Some packages may not be found! No local packages or download links found for python-Levenshtein error: Could not find suitable distribution for Requirement.parse('python-Levenshtein') Come on! Can you please read the error messages more carefully? It is pretty obvious it cannot connect to pypi.python.org in this case. That is also not a Scientific Linux related server. -- kind regards, David Sommerseth
Re: Anyone know of a status site for the Internet?
On 16/04/13 06:45, Mihamina Rakotomandimby wrote: On 2013-04-15 15:15, Paul Robert Marino wrote: In direct answer to your question oddly enough twitter seems to be the site find out about this kind of thing quickly. Twitter yes, but what to follow? Not necessarily follow ... ask on twitter and provide some relevant hashtags, like #scientificlinux -- kind regards, David Sommerseth
Re: SL 6.3
On 31/01/13 11:46, Arun Kishore wrote: Dear SL, I find updated repositories like below - good enough. i have been using below link for SL 6.3. I think it may be of help to you and all. There are many such similar sites.. for updates http://mirror.fraunhofer.de/download.fedora.redhat.com/epel/6/ This is just the EPEL repository ... I'd rather recommend just installing the yum-conf-epel package (yum install it), and you'll get a package which will be maintained with the always proper repository URLs - including mirrors. kind regards, David Sommerseth
Re: New install boot problem -
On 29/01/13 20:40, Trenton Ray wrote: GRUB has a strange methodology (well, strange to me given that I'm one of a handful of neo-luddites still using LILO) for determining disks. It's got root as hd2,0 which means (and I may be mistaken) the 2nd disk, first partition.. which makes sense in accordance with /dev/sdb but since GRUB sees 0 as a value and not a null, maybe change it to 1,0 ? I fear I'm just going to add to the confusing at this point and apologize for not being able to help offhand. Most commonly you'll find that: /dev/sda - hd0 /dev/sdb - hd1 /dev/sdc - hd2 And then with partitions: /dev/sdX1 - hdX,0 /dev/sdX2 - hdX,1 /dev/sdX3 - hdX,2 ... and so forth ... [root@box7 bobg]# cat /boot/grub/grub.conf # grub.conf generated by anaconda # # Note that you do not have to rerun grub after making changes to this file # NOTICE: You have a /boot partition. This means that # all kernel and initrd paths are relative to /boot/, eg. # root (hd2,0) # kernel /vmlinuz-version ro root=/dev/sdc3 # initrd /initrd-[generic-]version.img #boot=/dev/sdb This line ^^^ can set to /dev/sda if you want GRUB on your MBR of /dev/sda. But remove the leading # default=0 timeout=15 splashimage=(hd2,0)/grub/splash.xpm.gz This looks for the splashimage (background graphics) from /dev/sdb1 ... inside the grub/ directory on this partition. If you mount /dev/sdb1 on /boot ... that means /boot/grub on your file system. hiddenmenu title Scientific Linux (2.6.32-279.19.1.el6.x86_64) root (hd2,0) This expects kernels to be found on /dev/sdb1 kernel /vmlinuz-2.6.32-279.19.1.el6.x86_64 ro root=UUID=088eabf7-7644-4acb-9017-bc510456357b rd_NO_LUKS rd_NO_LVM LANG=en_US.UTF-8 rd_NO_MD SYSFONT=latarcyrheb-sun16 crashkernel=auto KEYBOARDTYPE=pc KEYTABLE=us rd_NO_DM rhgb quiet initrd /initramfs-2.6.32-279.19.1.el6.x86_64.img title Scientific Linux (2.6.32-279.5.1.el6.x86_64) root (hd2,0) Same here as well. kernel /vmlinuz-2.6.32-279.5.1.el6.x86_64 ro root=UUID=088eabf7-7644-4acb-9017-bc510456357b rd_NO_LUKS rd_NO_LVM LANG=en_US.UTF-8 rd_NO_MD SYSFONT=latarcyrheb-sun16 crashkernel=auto KEYBOARDTYPE=pc KEYTABLE=us rd_NO_DM rhgb quiet initrd /initramfs-2.6.32-279.5.1.el6.x86_64.img title Other rootnoverify (hd0,0) chainloader +1 This would typically load Windows or so, installed on /dev/sda1 Hope this helps you understand the drive references a bit better. -- kind regards, David Sommerseth
Re: SL-6.3 Install xfce -
On 29/01/13 21:42, Alec T. Habig wrote: Bob Goodwin - Zuni, Virginia, USA writes: I would like to install xfce but there doesn't seem to be a yum rpm. What is the best way to deal with that? It's there, but in the epel repository. Go get the epel-release package from here: http://fedoraproject.org/wiki/EPEL Or just do: yum install yum-conf-epel And you'll get all the needed files for EPEL ... no need to go via the Fedora wiki any more :) Then just do: yum groupinstall xfce That should be it. -- kind regards David Sommerseth
Re: SL-6.3 Install xfce -
On 29/01/13 22:47, Bob Goodwin - Zuni, Virginia, USA wrote: On 01/29/2013 03:57 PM, David Sommerseth wrote: On 29/01/13 21:42, Alec T. Habig wrote: Bob Goodwin - Zuni, Virginia, USA writes: I would like to install xfce but there doesn't seem to be a yum rpm. What is the best way to deal with that? It's there, but in the epel repository. Go get the epel-release package from here: http://fedoraproject.org/wiki/EPEL Or just do: yum install yum-conf-epel And you'll get all the needed files for EPEL ... no need to go via the Fedora wiki any more :) Then just do: yum groupinstall xfce That should be it. Yes, thanks, that eased the pain a lot! Now if I can figure out how to shange from gnome to xfce. Maybe there's a switchdesk rpm? When you log in via the graphical login screen, you should be able to choose desktop environment directly there. Usually at the bottom of the screen, iirc. -- kind regards, David Sommerseth
Re: SL-6 KVM Disk Activity
On 22/01/13 07:34, Chuck Munro wrote: - I tried to use the Virtio flavor of disk, but the BSD-based VMs can't use them, so I had to stay with IDE disk emulation. Depending on the BSD flavour, of course, but FreeBSD 8.2+ and 9 have virtio support ... http://www.area536.com/projects/freebsd-as-a-kvm-guest-using-virtio/ http://people.freebsd.org/~kuriyama/virtio/ I believe there are some ongoing work for virtio support in both OpenBSD and NetBSD as well. But that's also all I know. -- kind regards, David Sommerseth
Re: growisofs and dual disk (+R) failure
On 02/01/13 06:27, Andrew Z wrote: more details: it appears that cdrkit/wodim are pretty old and might not support well the DL burning. And indead, the genisoimage is a leftover from wodim. Would appreciate a (rpm) spec file for cdrtools. The one for suse is giving me hard time. As a normal (non-root) user, you can do this: $ yumdownloader --source cdrkit $ rpm -i cdrkit-*.el6.src.rpm $ cd ~/rpmbuild/SPECS And you'll have the cdrkit.spec file there. kind regards, David Sommerseth
Re: Does anyone have virt-manager working with sudo?
On 20/12/12 19:49, Paul Robert Marino wrote: Its base64 with DIGEST-MD5 hashing with no salt. If you don't beleave me just decode it through any base64 tool and you will see the entire conversation And if you still don't beleave me read the RFC that describes SASL its very clearly explained and a relativly short read as RFCs go. I've read through RFC2831 [1] more times now, which describes the DIGEST-MD5 protocol pretty well. And there are some details there, which libvirt user and which makes it impossible to use any base64 tool to extract the password, as you claim. First, this is the dialogue I captured between client and server: S: DIGEST-MD5 C: DIGEST-MD5 S: nonce=2VE5+WoJF+WSOHai6UpvyPUyjPV3TVdU//x25Bivx/w=, realm=virtsrv1.local, qop=auth-conf,cipher=rc4-56,rc4,3des, maxbuf=65536,charset=utf-8, algorithm=md5-sess C: username=vmadmin,realm=virtsrv1.local, nonce=2VE5+WoJF+WSOHai6UpvyPUyjPV3TVdU//x25Bivx/w=, cnonce=3G5mp5dUDXKs6Z/D9pvMLHMk177+Y20DNWfXoK0iscc=,nc=0001, qop=auth-conf,cipher=rc4, maxbuf=65536, digest-uri=libvirt/172.16.21.10, response=145b633fe038af33327d720bdb9111b0 S: rspauth=d0c38ff1236b11aa22966cafc78d67f8 This matches very well the protocol described in RFC2831. And there are a few things to notice here. a) qop-options From the RFC: A quoted string of one or more tokens indicating the quality of protection values supported by the server. The value auth indicates authentication; the value auth-int indicates authentication with integrity protection; the value auth-conf indicates authentication with integrity protection and encryption. This directive is optional; if not present it defaults to auth. The client MUST ignore unrecognized options; if the client recognizes no option, it should abort the authentication exchange. The server and clients both sends: qop=auth-conf b) cipher-opts From the RFC A list of ciphers that the server supports. This directive must be present exactly once if auth-conf is offered in the qop-options directive, in which case the 3des and des modes are mandatory-to-implement. The client MUST ignore unrecognized options; if the client recognizes no option, it should abort the authentication exchange. The server sends cipher=rc4-56,rc4,3des and the client responds with cipher=rc4. Which means a 128 bit RC4 encryption was chosen. c) response field from client From the RFC: A string of 32 hex digits computed as defined below, which proves that the user knows a password. This directive is required and MUST be present exactly once; otherwise, authentication fails. Further, when decoding the clients 'response', it is not representing anything 'text readable'. You can try to base64 or hex decode it. All you get is binary data. This field is built up by hex-encoding a concatenation of several fields, some which have been through a hashing round first. The first step is to hash this string: HASH1 = MD5($USERNAME:$REALM:$PASSWD) And then this hash is concatenated and hashed once more like this: HASH2 = MD5($HASH1:$NONCE:$CNONCE) $NONCE is the nonce value from the server, and $CNONCE is the cnonce value from the client. Both which are unique for this authentication session. An MD5 hashing algorithm is used on this string. A second string is prepared as well, like this: STR = MD5(AUTHENTICATE:$DIGESTURI:) The string 'AUTHENTICATE:' is added in front of $DIGESTURI which is the same as the digest-uri from the client. Then the final result is put together, using hex encoding of the input data: HEX( HEX($HASH2):$NONCE:$NCVALUE:$CNONCE:$QOP:HEX($STR) ) The $NONCE and $CNONCE are the same as in the previous step, but the $NCVALUE is the nc field from the client and $QOP is the qop field from the client. In addition, as the client responded with 'cipher=rc4', an encryption performed as well, with the key based on parts of $HASH2. But I will not go into the deep details here now. So if you still claim that libvirt sends passwords in clear-text, just BASE64 encoded, well, then I challenge you to expose the password in the client/server dialogue pasted above. How I understand this, the password has been hashed several times, with different salts (even though not proper random salts). And it is being RC4 encrypted in addition. I simply cannot find anything which looks like a BASE64 encoded password in neither the client/server dialogue nor the RFC - when qop is set to 'auth-conf' and having 'cipher' set. kind regards, David Sommerseth [1] RFC 2831 http://www.ietf.org/rfc/rfc2831.txt On Dec 20, 2012 10:16 AM, David Sommerseth sl+us...@lists.topphemmelig.net mailto:sl%2bus...@lists.topphemmelig.net wrote: On 19/12/12 23:45, Paul Robert
Re: Does anyone have virt-manager working with sudo?
On 19/12/12 23:45, Paul Robert Marino wrote: this statement (I double checked the network traffic, and even though not using SSL the password is not transferred over the network in clear-text) is wrong It is clear text that has been base64 encoded unless you are using gssapi with kerberos 5 or some other encrypted authentication mechanism. you can mitigate this slightly by using the digest-md5 auth but all you are doing then is sending the hash of the password in clear text which isn't that much better I might be wrong, but I am quite confident that the data I see is encrypted. I've looked at the data via wireshark, and even though it looks like base64, it doesn't render my password. And looking further in the comments in /etc/libvirt/libvirtd.conf, there is this statement: # Using the TCP socket requires SASL authentication by default. Only # SASL mechanisms which support data encryption are allowed. This is # DIGEST_MD5 and GSSAPI (Kerberos5) And what I see, looks pretty much like digest-md5 with RC4 encryption. My libvirtd daemon tells this to my client: DIGEST-MD5 Dnonce=BASE64_ecncoded_data, realm=DOMAIN,qop=auth-conf, cipher=rc4-56,rc4,3des, maxbuf=65536,charset=utf-8,algorithm=md5-sess The client responds with this: DIGEST-MD5 Eusername=USERNAME,realm=DOMAIN, nonce=BASE64_encoded_data, cnonce=BASE64_encoded_data_2, nc=0001,qop=auth-conf,cipher=rc4,maxbuf=65536, digest-uri=libvirt/1035110, response=1f4023d0417acb495ed187255ba80fcf And then the server responds with: rspauth=40922952d194910a08f3654b28d5485f After this response, it comes a data flow which looks more like normal data traffic. As far as I can interpret this, the data from the client is encrypted, using some shared data for a key exchange. But I haven't dug into the source code to verify this yet. If I'm wrong, then I need to learn more about the DIGEST-MD5 protocol. kind regards, David Sommerseth On Wed, Dec 19, 2012 at 6:14 AM, David Sommerseth sl+us...@lists.topphemmelig.net wrote: On 19/12/12 04:38, Nico Kadel-Garcia wrote: I'd love to be able to use sudo with virt-manager, but it simply fails. It does work on Ubuntu, and I'd like to be able to use sudo for all access to my KVM servers, rather than direct root login. Is anyone using sudo successfully with virt-manager on SL 6.3? Other X applications work just fine. I do something similar to this, but slightly differently. I'm running virt-manager as my local user, but connecting to the TCP port, authenticating using the SASL feature. (I double checked the network traffic, and even though not using SSL the password is not transferred over the network in clear-text) * In /etc/libvirt/libvirtd.conf I set the following parameters: listen_tls=0 listen_tcp=1 listen_address=x.x.x.x # optional auth_tcp = sasl * Create the SASL database with the username/password I want to use Look up the sasldb_path in /etc/sasl2/libvirt.conf. In my setup it was set to /etc/libvirt/passwd.db. Then do: [root@host ~]# saslpasswd2 -f /etc/libvirt/passwd.db USERNAME The username can be completely virtual if you want. saslpasswd2 will ask for the wanted password. This username/password will only be used for libvirt. You can check the user database like this: [root@host ~]# sasldblistusers2 -f /etc/libvirt/passwd.db * Make libvirtd listen to TCP sockets Edit /etc/sysconfig/libvirtd so that LIBVIRTD_ARGS have --listen set F.ex: LIBVIRTD_ARGS=--listen * Restart libvirtd [root@host ~]# service libvirtd restart If you want to connect to another IP than localhost, you also need to open up your firewall as well. But at this point, it should be possible to connect to libvirt over the network: [user@host ~]$ virsh -c qemu+tcp://localhost/system [user@host ~]$ virt-manager -c qemu+tcp://localhost/system I'm primarily using this approach to manage a couple of KVM hosts over a VPN connection, running virsh and virt-manager locally, connecting to the IP which is accessible over the VPN. Works pretty well, but I had to do some changes in the guest configs so that SPICE or VNC consoles also listened to an IP address available over the VPN. Changing these parameters requires cold-booting the virtual guests. I have tried to use SSL/TLS mode as well, but when managing more independent libvirt based servers, it gets quite annoying with how the client certificates needs to be configured. (Granted, it's a long time ago I tried it, so it might have changed) But I figured, as long as I use SASL and does this over VPN connections with strict firewalls, the setup is safe enough in my use cases. kind regards, David Sommerseth
Re: Does anyone have virt-manager working with sudo?
On 19/12/12 04:38, Nico Kadel-Garcia wrote: I'd love to be able to use sudo with virt-manager, but it simply fails. It does work on Ubuntu, and I'd like to be able to use sudo for all access to my KVM servers, rather than direct root login. Is anyone using sudo successfully with virt-manager on SL 6.3? Other X applications work just fine. I do something similar to this, but slightly differently. I'm running virt-manager as my local user, but connecting to the TCP port, authenticating using the SASL feature. (I double checked the network traffic, and even though not using SSL the password is not transferred over the network in clear-text) * In /etc/libvirt/libvirtd.conf I set the following parameters: listen_tls=0 listen_tcp=1 listen_address=x.x.x.x # optional auth_tcp = sasl * Create the SASL database with the username/password I want to use Look up the sasldb_path in /etc/sasl2/libvirt.conf. In my setup it was set to /etc/libvirt/passwd.db. Then do: [root@host ~]# saslpasswd2 -f /etc/libvirt/passwd.db USERNAME The username can be completely virtual if you want. saslpasswd2 will ask for the wanted password. This username/password will only be used for libvirt. You can check the user database like this: [root@host ~]# sasldblistusers2 -f /etc/libvirt/passwd.db * Make libvirtd listen to TCP sockets Edit /etc/sysconfig/libvirtd so that LIBVIRTD_ARGS have --listen set F.ex: LIBVIRTD_ARGS=--listen * Restart libvirtd [root@host ~]# service libvirtd restart If you want to connect to another IP than localhost, you also need to open up your firewall as well. But at this point, it should be possible to connect to libvirt over the network: [user@host ~]$ virsh -c qemu+tcp://localhost/system [user@host ~]$ virt-manager -c qemu+tcp://localhost/system I'm primarily using this approach to manage a couple of KVM hosts over a VPN connection, running virsh and virt-manager locally, connecting to the IP which is accessible over the VPN. Works pretty well, but I had to do some changes in the guest configs so that SPICE or VNC consoles also listened to an IP address available over the VPN. Changing these parameters requires cold-booting the virtual guests. I have tried to use SSL/TLS mode as well, but when managing more independent libvirt based servers, it gets quite annoying with how the client certificates needs to be configured. (Granted, it's a long time ago I tried it, so it might have changed) But I figured, as long as I use SASL and does this over VPN connections with strict firewalls, the setup is safe enough in my use cases. kind regards, David Sommerseth
Re: Desktop Cinnamon
On 13/12/12 14:32, Larry Linder wrote: Has anyone tried to use Cinnamon desktop onto SL. The Desktops on 6.X were unusable from our vantage point and that is why we stalled at 5.8. From some others I have talked to have tried it and found that they love it and they indicated that were running Fedora. Cinnamon requires GNOME 3. It's basically just a GNOME Shell replacement in GNOME 3. So to run Cinnamon on EL6, you need to bring over the complete GNOME 3 stack as well. I know Cinnamon is available in Fedora 17. kind regards, David Sommerseth
Re: Cannot start luci
On 12/12/12 17:28, Evan Sather wrote: Hi Oleg, Here is what repeats in my luci.log: Traceback (most recent call last): File /usr/bin/paster, line 9, in module load_entry_point('PasteScript==1.7.3', 'console_scripts', 'paster')() File /usr/lib/python2.6/site-packages/paste/script/command.py, line 84, in run invoke(command, command_name, options, args[1:]) File /usr/lib/python2.6/site-packages/paste/script/command.py, line 123, in invoke exit_code = runner.run(args) File /usr/lib/python2.6/site-packages/paste/script/command.py, line 218, in run result = self.command() File /usr/lib/python2.6/site-packages/paste/script/serve.py, line 274, in command relative_to=base, global_conf=vars) File /usr/lib/python2.6/site-packages/paste/script/serve.py, line 308, in loadserver relative_to=relative_to, **kw) File /usr/lib/python2.6/site-packages/paste/deploy/loadwsgi.py, line 210, in loadserver return loadobj(SERVER, uri, name=name, **kw) File /usr/lib/python2.6/site-packages/paste/deploy/loadwsgi.py, line 224, in loadobj global_conf=global_conf) File /usr/lib/python2.6/site-packages/paste/deploy/loadwsgi.py, line 248, in loadcontext global_conf=global_conf) File /usr/lib/python2.6/site-packages/paste/deploy/loadwsgi.py, line 278, in _loadconfig return loader.get_context(object_type, name, global_conf) File /usr/lib/python2.6/site-packages/paste/deploy/loadwsgi.py, line 363, in get_context object_type, name=name) File /usr/lib/python2.6/site-packages/paste/deploy/loadwsgi.py, line 528, in find_config_section self.filename)) LookupError: No section 'init' (prefixed by 'server') found in config /var/lib/luci/etc/luci.ini Removing PID file /var/run/luci/luci.pid When I installed luci, I had to do a touch /var/lib/luci/etc/luci.ini because I couldn't get past this start failure message: -bash-4.1$ sudo service luci start Unable to create the luci base configuration file (`/var/lib/luci/etc/luci.ini'). Start luci... [FAILED] Did I miss a step after installing luci that I should have done before trying to start it? Yes, you missed reading the error message: No section 'init' (prefixed by 'server') found in config /var/lib/luci/etc/luci.ini I'm no luci user so can't help you much there. But looks like you need to get a proper configuration file set up, not just use touch and hope it works ;-) kind regards, David Sommerseth
Re: Scala missing in the official repo(s)
On 29/11/12 13:14, Freak Trick wrote: The developers at Scala themselves offer an RPM in the downloads section of the official website, but it conflicts with the jline.jar file of the JVM. I have never done/maintained packages. Let me check out some resources over the net and see if I can make a useful one. If I can, I'd definately look to contribute it. Why make it so difficult? ;-) I would go first go here: http://koji.fedoraproject.org/koji/packageinfo?packageID=6830 If I find.el{5,6} packages here, I would go and install the Fedora EPEL repository (yum install yum-conf-epel, iirc). If there are no EPEL packages available, I would find the appropriate package version and probably choose the builds for oldest Fedora releases (EL6 is based on Fedora 12/Fedora 13) and download the .src.rpm. Then I would install mock and rpmbuild (yum install mock rpm-build) ... then you just do: $ mock -r epel-6-x86_64 --rebuild /path/to/scala-*.src.rpm (replace x86_64 with i386 to get a 32bit build ... for other targets, see the files in /etc/mock) After a little while, you'll get what you need in /var/lib/mock ... ready to be installed with 'yum localinstall rpms' Using mock will download and install all the needed stuff for compiling the src.rpm in a chroot in /var/lib/mock ... so you'll need some space there to be able to make this work. But the log files in this directory are usually quite helpful as well. -- kind regards, David Sommerseth
Re: SL6.3 - groupinstall error
On 15/11/12 17:33, Duke Nguyen wrote: On 11/15/12 11:27 PM, David Sommerseth wrote: On 15/11/12 16:56, Duke Nguyen wrote: On 11/15/12 7:17 PM, David Sommerseth wrote: On 15/11/12 12:03, Duke Nguyen wrote: Hi folks, I am trying to install a Base system into a directory using yum groupinstall, but I got error as below. Any suggestion to solve the errors? Thanks. $ cat /etc/redhat-release Scientific Linux release 6.3 (Carbon) $ uname -a Linux biowulf.grih.org 2.6.32-279.14.1.el6.x86_64 #1 SMP Tue Nov 6 11:21:14 CST 2012 x86_64 x86_64 x86_64 GNU/Linux $ sudo yum -y groupinstall Base Server Platform --installroot=/diskless/root Loaded plugins: fastestmirror, refresh-packagekit, security Loading mirror speeds from cached hostfile * sl: ftp.scientificlinux.org * sl-security: ftp.scientificlinux.org http://ftp.scientificlinux.org/linux/scientific/%24releasever/x86_64/os/repodata/repomd.xml: ^^^ This looks wrong. This should be $releasever and should have been expanded to 6.3 before sent to the web server. Not sure why this happens though :/ %24 is ASCII hex code for $. Thanks David, I also noticed that. But that is only one strang error! I did try to change $releasever to 6.3 in all repo files in /etc/yum.repo.d/, and did not have above error, but still got similar error at the end: -- Processing Dependency: kernel = 2.6.9-11 for package: systemtap-runtime-1.7-5.el6.x86_64 --- Package xml-common.noarch 0:0.6.3-32.el6 will be installed -- Finished Dependency Resolution Error: Package: libdrm-2.4.25-2.el6.x86_64 (sl) Requires: kernel = 2.6.29.1-52.fc11 Error: Package: pcmciautils-015-4.2.el6.x86_64 (sl) Requires: kernel = 2.6.12-1.1411_FC5 Error: Package: systemtap-runtime-1.7-5.el6.x86_64 (sl) Requires: kernel = 2.6.9-11 You could try using --skip-broken to work around the problem You could try running: rpm -Va --nofiles --nodigest Question is: why the system installed fine with live CD, and now with the exact same system, installation to different location (/diskless/root) fails with all kind of dependencies? I'll admit I didn't look too carefully at the rest last time. Fixing the first steps usually helps solving the next ones ... but there are a couple of things here I wonder about: kernel = 2.6.29.1-52.fc11 kernel = 2.6.12-1.1411_FC5 kernel = 2.6.9-11 That is ancient kernels, none of them have (to my knowledge) ever been EL kernels, and all of them predates EL6 and even EL5. So I'm wondering where it got the information about these kernels. Could you please provide the output of 'rpm -qa kernel' ? Here you go: $ rpm -qa kernel kernel-2.6.32-279.14.1.el6.x86_64 kernel-2.6.32-279.5.1.el6.x86_64 Hmm. Just a silly question ... but have you tried a 'yum clean all' and then tried the groupinstall? It's really peculiar that it complains about these old kernels when you don't have them - and the versions you have should really be good enough. kind regards, David Sommerseth
Re: SL6.3 - groupinstall error
On 15/11/12 12:03, Duke Nguyen wrote: Hi folks, I am trying to install a Base system into a directory using yum groupinstall, but I got error as below. Any suggestion to solve the errors? Thanks. $ cat /etc/redhat-release Scientific Linux release 6.3 (Carbon) $ uname -a Linux biowulf.grih.org 2.6.32-279.14.1.el6.x86_64 #1 SMP Tue Nov 6 11:21:14 CST 2012 x86_64 x86_64 x86_64 GNU/Linux $ sudo yum -y groupinstall Base Server Platform --installroot=/diskless/root Loaded plugins: fastestmirror, refresh-packagekit, security Loading mirror speeds from cached hostfile * sl: ftp.scientificlinux.org * sl-security: ftp.scientificlinux.org http://ftp.scientificlinux.org/linux/scientific/%24releasever/x86_64/os/repodata/repomd.xml: ^^^ This looks wrong. This should be $releasever and should have been expanded to 6.3 before sent to the web server. Not sure why this happens though :/ %24 is ASCII hex code for $. kind regards, David Sommerseth
Re: SL6.3 - groupinstall error
On 15/11/12 16:56, Duke Nguyen wrote: On 11/15/12 7:17 PM, David Sommerseth wrote: On 15/11/12 12:03, Duke Nguyen wrote: Hi folks, I am trying to install a Base system into a directory using yum groupinstall, but I got error as below. Any suggestion to solve the errors? Thanks. $ cat /etc/redhat-release Scientific Linux release 6.3 (Carbon) $ uname -a Linux biowulf.grih.org 2.6.32-279.14.1.el6.x86_64 #1 SMP Tue Nov 6 11:21:14 CST 2012 x86_64 x86_64 x86_64 GNU/Linux $ sudo yum -y groupinstall Base Server Platform --installroot=/diskless/root Loaded plugins: fastestmirror, refresh-packagekit, security Loading mirror speeds from cached hostfile * sl: ftp.scientificlinux.org * sl-security: ftp.scientificlinux.org http://ftp.scientificlinux.org/linux/scientific/%24releasever/x86_64/os/repodata/repomd.xml: ^^^ This looks wrong. This should be $releasever and should have been expanded to 6.3 before sent to the web server. Not sure why this happens though :/ %24 is ASCII hex code for $. Thanks David, I also noticed that. But that is only one strang error! I did try to change $releasever to 6.3 in all repo files in /etc/yum.repo.d/, and did not have above error, but still got similar error at the end: -- Processing Dependency: kernel = 2.6.9-11 for package: systemtap-runtime-1.7-5.el6.x86_64 --- Package xml-common.noarch 0:0.6.3-32.el6 will be installed -- Finished Dependency Resolution Error: Package: libdrm-2.4.25-2.el6.x86_64 (sl) Requires: kernel = 2.6.29.1-52.fc11 Error: Package: pcmciautils-015-4.2.el6.x86_64 (sl) Requires: kernel = 2.6.12-1.1411_FC5 Error: Package: systemtap-runtime-1.7-5.el6.x86_64 (sl) Requires: kernel = 2.6.9-11 You could try using --skip-broken to work around the problem You could try running: rpm -Va --nofiles --nodigest Question is: why the system installed fine with live CD, and now with the exact same system, installation to different location (/diskless/root) fails with all kind of dependencies? I'll admit I didn't look too carefully at the rest last time. Fixing the first steps usually helps solving the next ones ... but there are a couple of things here I wonder about: kernel = 2.6.29.1-52.fc11 kernel = 2.6.12-1.1411_FC5 kernel = 2.6.9-11 That is ancient kernels, none of them have (to my knowledge) ever been EL kernels, and all of them predates EL6 and even EL5. So I'm wondering where it got the information about these kernels. Could you please provide the output of 'rpm -qa kernel' ? kind regards, David Sommerseth
Re: flash replacement
On 14/11/12 10:58, Todd And Margo Chester wrote: On 11/13/2012 08:57 PM, Andrew Z wrote: hello, is there an alternative for ( lots of grumpy swearing omitted ) flash plugin in FF? with the latest 11.2.202.251 x86 release the HD videos in full-screen a crashing. [bitching on] first i deal with an invasion of Avatars - the blue people and now $$$ thing won't even play for longer than 4 minutes in full screen [bitching off] Hi Andrew, Oh good. I get to help someone for a change. The recent Flash player is a DISASTER. Can't seem to get anyone to fix it either. Back off to flash-plugin-11.1.102.63-0.1.el6.rf.x86_64.rpm (or 32 bit, if that's what you are running) and turn off auto updates. http://apt.sw.be/redhat/el6/en/x86_64/dag/RPMS/flash-plugin-11.1.102.63-0.1.el6.rf.x86_64.rpm # rpm -e flash-plugin # rpm -ivh flash-plugin-11.1.102.63-0.1.el6.rf.x86_64.rpm Seriously, what's wrong with using the official Adobe Flash player? I see you're using one from the rpmforge/fusion repo. 1. Go here: http://get.adobe.com/flashplayer/ 2. Select YUM for Linux and click download. 3. Install that RPM 4. as root: yum erase flash-* 5. Disable all those extra elrepo, atrpms, rpmfusion/rpmforge repositories (look inside the files in /etc/yum.repos.d/ ) 5. as root: yum install flash-plugin 6. Enjoy a more reliable flash. This way you get no 32bit nonsense on 64bit installations. And you get something which works quite more reliable. I'm using the Adobe Flash player (64bit) on a couple of my private laptops. And I'm watching fullscreen TV broadcasts in hi quality resolution (3.5Mbit/s streams) scaled up to my 1920x1200 display ... and if there are issues, that's been related to not getting the needed bandwidth. I've mostly had just troubles with those additional repos, so I only enable them when it's needed. Otherwise, SL repos + EPEL repos gives a really stable and reliable systems At least, that's been my experience. kind regards, David Sommerseth
Re: recent kernel and root raid1
On 12/11/12 22:14, Konstantin Olchanski wrote: On Sat, Nov 10, 2012 at 08:14:41AM -0600, Robert Blair wrote: I have a system that failed to boot after the most recent kernel update. It took a while, but I eventually traced it to the initramfs not having raid1 included. I had to manually do a mkinitrd --preload raid1 ... ... dracut-004-283.el6.noarch All are invited to inspect the mdadm-related code in dracut. Instead of running mdadm -As and hoping for the best, there is a maze of twisty write-only shell scripts to do who knows what (there is no documentation whatsoever) and they seem to get things wrong as often as they get them right. People who develop that kind of stuff and push it on us as improvements should have their heads inspected. (Insult intended). I've been uncertain if I should respond to this mail or not. But I struggle to let such rants pass when I think this is a really an unfair attack. I will not respond to the technical solutions chosen by upstream communities or projects (mdadm, dracut, etc). But I do react to your claim which sounds like you think the upstream developers are clueless and don't care about what they do. In addition you do this on a mailing list which is not directly related to the upstream components which you criticise. All the code in this Enterprise Linux distro comes from an open source upstream source. There are usually upstream communities to discuss things with. And there are communities where you can provide patches fixing things you find not being in a good shape. Insightful feedback is always appreciated. Using the proper channels to provide feedback and patches can at least result in something fruitful. Like improving dracut. https://dracut.wiki.kernel.org/index.php/Main_Page Throwing out such trash which you did will definitely not improve anything. Upstream developers do really deserve better treatment from us - no matter what we think of their work. And especially if they don't hang around on this mailing list to respond to such (intended) insults. So please, can we try to lift these discussions from throwing dirt behind peoples back to rather provide useful feedback on the right places? Thank you! kind regards, David Sommerseth
Re: recent kernel and root raid1
On 13/11/12 21:04, Konstantin Olchanski wrote: On Tue, Nov 13, 2012 at 08:33:05PM +0100, David Sommerseth wrote: On 12/11/12 22:14, Konstantin Olchanski wrote: ... But I do react to your claim which sounds like you think the upstream developers are clueless and don't care about what they do. That is not what I said. I did not say they are stupid, I did not say they are indifferent or malicious. I said that they are crazy. There is a difference. You did not use the term crazy, but you did indicate very well you wanted to insult someone. Insults in general are not something I personally appreciate, and feel rather confident I'm not alone in that opinion. No matter how it is expressed. All the code in this Enterprise Linux distro comes from an open source upstream source. There are usually upstream communities to discuss things with. And there are communities where you can provide patches fixing things you find not being in a good shape. There is a minor problem with your suggestion that I send bug fixes to upstream: Upstream is at dracut-024 (Oct-2012), SL6 is at dracut-004 (January-2010). Yes, you are right. But you are also wrong. Yes, SL6 is based on the dracut-004 release. But that does not mean the latest SL6 dracut package is comparable to the -004 release any more. There has been quite some fixes since that release. Just do a 'rpm -q --changelog dracut' and see for yourself. Even better, download the src.rpm and look at the patches applied on top of the -004 release. Just an example. The latest change to dracut-004-284 contains 0284-fips-set-boot-as-symlink-to-sysroot-boot-if-no-boot-.patch. This points at upstream git commit f22f08d857fb239c38daf7099fc55e82506f4abe which can be found in the RHEL-6 branch in the upstream git tree. Important fixes are included and fixed when needed. So if it doesn't work for you, get in touch with upstream and/or file a bugzilla with upstream project. If it is important for RHEL, it will arrive into RHEL at some point too. The severity of the issue decides how quickly a fix gets pushed out. And when RHEL pushes out a fix ... then SL gets the fix too. Bottom line is: Working together with upstream projects does benefit SL too. Red Hat does a lot of backports. A package is frozen on a specific version to be stabilised when the next RHEL release is being worked on. From that point of, the idea is that fixes and enhancements are backported from newer versions. As an example. KVM was introduced in the 2.6.20 kernel. RHEL5 ships with a 2.6.18 based kernel. How come RHEL5 supports KVM? I can elaborate on how bug fixes to upstream do no good to SL users. No need. Throwing out such trash which you did will definitely not improve anything. Upstream developers do really deserve better treatment from us - no matter what we think of their work. Please refer to: http://en.wikipedia.org/wiki/The_Emperor's_New_Clothes I'm very well aware of that fairytale, but I'm not sure why you bring that one up in this discussion. My point is that telling the right people upfront of issues generally makes things better. Ranting behind peoples backs sounds more familiar to the fairytale to me. kind regards, David Sommerseth
Re: mt and LTO
On 01/11/12 22:57, Jeff Siddall wrote: Not to bash tools like cpio, tar and mt, but one of the coolest pieces of software out there has to be BackupPC. Huge cheap magnetic disks combined with the compression and file pooling capabilities of BackupPC makes it hard for me to imagine going back to tapes. Just last week I restored an entire 1.5 GB LTSP client image from a backup made a few months ago, all in a matter of a few minutes with no digging for tapes. I've been using BoxBackup for quite some time. This is an online backup solution. It might not be as fancy as Amanda or Bacula, but it is darn easy to setup and works quite well. Most of my boxes are setup with lazy backup, which basically means rolling backups based time stamps on the files. All backups are encrypted on the client side too, before they are sent to the server. This means you'll have to have the decryption key saved somewhere else safely to be able to do disaster recoveries. But this is can be ideal for off-site storage where you might not trust the storage environment. What I don't like about BoxBackup is that the project almost seems dead, even though some developement is happening. And the command line tools can sometimes be a bit tricky to work with. But when you find the right tool and the right commands, it seems to work very well. There are also some issues with deleted files being cleaned up a bit too early and something about timestamps on directories, iirc. The Windows client does not handle files over 2GB well, but those files can be restored on a Linux computer (provided you have the encryption key available) If I get some more time, I'm planning to submit BoxBackup to Fedora (including EPEL) after I've cleaned up the .spec file a bit more (the one in the source tree is rather nasty, and definitely does not comply with current Fedora packaging rules). But if someone is interested to give this a try and/or help out, get in touch and I'll provide a src.rpm for testing. But I have to admit ... if Bacula would have been as easy to set up (including encryption) as BoxBackup, I would probably have been using Bacula instead - mostly because of the tools and it seems to be better developed. Bacula can use file storage as well as tapes. Another option for backup to hard drives is also to use iSCSI and export partitions or files as iSCSI tape volumes. I've not tested this yet, but if the docs is correct, it should even support mtx to exchange tapes. kind regards, David Sommerseth
Re: [SCIENTIFIC-LINUX-USERS] Yum update problem -
(sorry! resending with proper e-mail address) On 31/10/12 15:50, Bob Goodwin - Zuni, Virginia, USA wrote: On 31/10/12 10:14, Pat Riehecky wrote: syslinux-4.05-1.el6.rfx.x86_64 is not from SL, I suspect your problem is coming from there. Pat Well where would it come from? I would have chosen an XFCE Live USB installation if that was available but I'm not certain ... Bob Try this: # yum --disablerepo=* --enablerepo=sl --enablerepo=sl-security update You might also consider to add --enablerepo=sl-fastbugs. With this approach, you first disable all repos and will only add the official SL repos. That should hopefully get you through the worst part. Otherwise, have a look in /etc/yum.repos.d ... and use rpm -qif repo-file to see where this file came from. If it's something you've added manually, you won't find a match. Otherwise, it's usually coming from a package somewhere. My experience is that rpmforge/rpmfusion and elrepo repositories often cause conflicts, so if I need them I have them disabled by default. This might be because I'm often depending on the EPEL repository too. YMMV. kind regards, David Sommerseth
Re: DeltaRPM Support for Scientific Linux
- Original Message - From: Piruthiviraj Natarajan piruthivi...@gmail.com To: SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV Cc: Piruthiviraj Natarajan piruthivi...@gmail.com Sent: Saturday, 13 October, 2012 8:34:41 PM Subject: DeltaRPM Support for Scientific Linux Hi Everyone, I 'm new to this list and Scientific Linux. I have been using Fedora and RHEL based clones for a while. I was thinking that it would a big benefit to the users to save some bandwidth if SL deployed Delta updates in the official repos. I asked the question in forum and they directed me here. where can I make the request for the feature? Just do: [root@host ~]# yum install yum-presto That's all, the presto plug-in is enabled automatically when installing it, and then delta-rpms are pulled down on updates. It works quite fine for me at least. kind regards, David Sommerseth
Re: DeltaRPM Support for Scientific Linux
- Original Message - From: Akemi Yagi amy...@gmail.com To: David Sommerseth sl+us...@lists.topphemmelig.net Cc: Piruthiviraj Natarajan piruthivi...@gmail.com, SCIENTIFIC-LINUX-USERS@listserv.fnal.gov Sent: Saturday, 13 October, 2012 10:08:49 PM Subject: Re: DeltaRPM Support for Scientific Linux On Sat, Oct 13, 2012 at 12:26 PM, David Sommerseth sl+us...@lists.topphemmelig.net wrote: - Original Message - From: Piruthiviraj Natarajan piruthivi...@gmail.com To: SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV Hi Everyone, I 'm new to this list and Scientific Linux. I have been using Fedora and RHEL based clones for a while. I was thinking that it would a big benefit to the users to save some bandwidth if SL deployed Delta updates in the official repos. I asked the question in forum and they directed me here. where can I make the request for the feature? Just do: [root@host ~]# yum install yum-presto That's all, the presto plug-in is enabled automatically when installing it, and then delta-rpms are pulled down on updates. It works quite fine for me at least. kind regards, David Sommerseth Are you sure about this? My understanding is that deltaRPMs are not available for SL. I know CentOS has them but ... Oh, good point! I might mix it with some other installations (with Fedora and RHEL). I know have installed presto on SL boxes as well, so I've taken it for granted. And it might be that I've seen it in action on third-party repos like EPEL or so. kind regards, David Sommerseth
Re: The opposite SL and VirtualBox problem
- Original Message - From: Nico Kadel-Garcia nka...@gmail.com To: David Sommerseth sl+us...@lists.topphemmelig.net Cc: Joseph Areeda newsre...@areeda.com, SCIENTIFIC-LINUX-USERS@listserv.fnal.gov, owner-scientific-linux-us...@listserv.fnal.gov Sent: Thursday, 4 October, 2012 2:53:01 AM Subject: Re: The opposite SL and VirtualBox problem On Tue, Oct 2, 2012 at 6:59 PM, David Sommerseth sl+us...@lists.topphemmelig.net wrote: - Original Message - From: Joseph Areeda newsre...@areeda.com To: owner-scientific-linux-us...@listserv.fnal.gov Cc: SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV Sent: Tuesday, 2 October, 2012 10:51:52 PM Subject: Re: The opposite SL and VirtualBox problem Well, I'm not going to touch Nico's comment because I don't know KVM. For me it's the Devil you know kind of thing. I've had good experience with Vbox on multiple OS and am just playing in my comfort zone. I do have reasons to explore other VMs but none of them pressing. I just want to install one of the University's free site license copy of Windows as a courtesy to our students. Even though Nico have some good points, I feel some of them are also dated due to the shape of virt-manager in earlier versions. In EL6.3, it's become quite good IMO and very usable. If you're running KVM locally on your own computer, there would be no benefits of using vbox IMO. What do you find improved? I'm writing a new KVM setup guideline for complete newbies on an open source project, and would welcome your insights. I did a 6.3 based installation today and found no significant improvementn in the virt-manager itself. I've been playing with virt-manager since Fedora 11 (maybe even a little bit in Fedora 8/9-ish) and RHEL5, so that's my background. And from that perspective, I can now easily through virt-manager create a new network bridge [1], setup LVM controlled storage pools (or adding iSCSI based disks [2], making the iSCSI layer invisible to the KVM guest) and create new VMs using the virt-manager configured network bridge and storage pool (including allocating a logical volume in the storage pool's volume group). I have all OSes downloaded as ISO images and saved them under /var/lib/libvirt/images and uses the ISO install from that directory. Or I can mount these ISO images and provide them via a locally configured http server and use the network install. And I do all this via the virt-manager. I've not played much with the Virtual Network feature, but from what I can see, it should be fairly straight forward to also configure that - as either a closed, NATed or routed network, where libvirt configures dnsmasq automatically if you want DHCP on the virtual network. To set up the network and storage pools, I right click on the localhost entry in virt-manager and click on Details. The VM installation is done using the Create new VM wizard, just filling out the fields. And VM management is done double clicking the VMs. So to answer your question, in the early days, all this was cumbersome and didn't always work. In SL6.3, I have no troubles doing this at all, It Just Works. So I struggle to understand your criticism of the virt-manager functionality (from a GUI perspective). It might be I'm blind to those issues you see, as I've had my fights earlier and which I don't have any more. So it almost feels like it is not doing it the way you want it to work, to which I can't add much. kind regards, David Sommerseth [1] If adding VLAN and bridges to the mix, you need to manually add VLAN=yes to the ifcfg-eth* file virt-manager creates of the eth device. Then creating the bridge afterwords will allow you to complete the setup without issues. That's the only issue I've hit in virt-manager in my last tests. [2] http://berrange.com/posts/2010/05/04/provisioning-kvm-virtual-machines-on-iscsi-with-qnap-virt-manager-part-2-of-2/
Re: Iptable rule required to block youtube
- Original Message - From: vivek chalotra vivekat...@gmail.com To: Henrique Junior henrique...@gmail.com Cc: Konstantin Olchanski olcha...@triumf.ca, scientific-linux-us...@fnal.gov Sent: Friday, 5 October, 2012 9:10:24 AM Subject: Re: Iptable rule required to block youtube I have blocked youtube(ips from 74.125.236.0- 74.125.236.14) in my gateway machine using the below rules: iptables -A INPUT -i eth1 -s 74.125.236.0 -j DROP iptables -A INPUT -i eth1 -p tcp -s 74.125.236.0 -j DROP iptables -A INPUT -i eth0 -s 74.125.236.0 -j DROP iptables -A INPUT -i eth0 -p tcp -s 74.125.236.0 -j DROP but how to block on the whole network. Other hosts are still able to access youtube. With whole network, do you mean your local LAN which your firewall (this SL box you're configuring) controls? If so, you should probably add those DROP rules to the FORWARD chain and not the INPUT chain. See this URL for more info: http://www.netfilter.org/documentation/HOWTO//packet-filtering-HOWTO-6.html kind regards, David Sommerseth
Re: The opposite SL and VirtualBox problem
- Original Message - From: Nico Kadel-Garcia nka...@gmail.com To: David Sommerseth sl+us...@lists.topphemmelig.net Cc: Joseph Areeda newsre...@areeda.com, SCIENTIFIC-LINUX-USERS@listserv.fnal.gov Sent: Tuesday, 2 October, 2012 1:53:29 PM Subject: Re: The opposite SL and VirtualBox problem On Tue, Oct 2, 2012 at 6:15 AM, David Sommerseth sl+us...@lists.topphemmelig.net wrote: [...snip...] When I am on a remote connection with limited bandwidth, and I need to add or modify hardware configurations for a VM, I *do not want* to have to run the graphics console of the VM simply in order to add a network port. The virt-manager tool got this *dead wrong*. Well, I might be quite used to work with low-level stuff. But I find 'virsh' quite handy to work with when just using ssh to my KVM host. And when I wonder about anything, this resource provides the info I need to figure out things: http://libvirt.org/sources/virshcmdref/html-single/ Otherwise, I've set up libvirt to be accessible over a TCP port, with authentication. So I use a local running instance of virt-manager, managing a remote KVM host. And it doesn't even really have to be the same version for general functionality. Even though, installing a new OS have been a bit annoying (unless doing networked installs) Oh, and did I forget to mention that if your X setup is not just right, virt-manager fails to start completely silently? It doesn't evem check if your DISPLAY is set and send an appropriate error message? Agreed, that's annoying. KVM also has very serious problems with network configuration. The necessary bridged network configuration for VM's that are not going to be behind a NAT or completely isolated is not actually supported by any of our upstream vendor's configuration tools. You have to hand edit your core network configuration files with a text editor, and any use of NetworkManager to manage VPN or wireless connections puts you at risk of breaking it unless you very rigorously and manually put 'NM_CONTROLLED=no' in each /etc/sysconfig/network-scripts/ifcfg-* file. It's a while since I tried to set up this. And last time, I needed VLAN's as well, so that required manually tweaking things even more. But I know it used to be like that. I haven't really tried it with the latest 6.3 release. It is a very, very steep learning curve to get your first KVM setups working, where with VirtualBox it's very plug and play. VirtualBox is unlikely to have the high scalability, live server migration, or kernel integration that KVM has, but for a casual virtual machine or two running common operating systems on common architectures, who cares? It used to be hard to setup KVM. Maybe I'm just gotten so used to the process, but I find it quite intuitive mostly these days. And the docs have also improved. Anyhow, I'm also quite sure the libvirt guys wouldn't mind patches from users fixing their itches as well ... as you mentioned you did some coding, I mean. kind regards, David Sommerseth
Re: The opposite SL and VirtualBox problem
- Original Message - From: Joseph Areeda newsre...@areeda.com To: owner-scientific-linux-us...@listserv.fnal.gov Cc: SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV Sent: Tuesday, 2 October, 2012 10:51:52 PM Subject: Re: The opposite SL and VirtualBox problem Well, I'm not going to touch Nico's comment because I don't know KVM. For me it's the Devil you know kind of thing. I've had good experience with Vbox on multiple OS and am just playing in my comfort zone. I do have reasons to explore other VMs but none of them pressing. I just want to install one of the University's free site license copy of Windows as a courtesy to our students. Even though Nico have some good points, I feel some of them are also dated due to the shape of virt-manager in earlier versions. In EL6.3, it's become quite good IMO and very usable. If you're running KVM locally on your own computer, there would be no benefits of using vbox IMO. You might find the RHEL6 Virtualisation guides quite handy though, Getting started: https://access.redhat.com/knowledge/docs/en-US/Red_Hat_Enterprise_Linux/6/html/Virtualization_Getting_Started_Guide/index.html Administration Guide: https://access.redhat.com/knowledge/docs/en-US/Red_Hat_Enterprise_Linux/6/html/Virtualization_Administration_Guide/index.html Host Configuration and Guest Installation Guide: https://access.redhat.com/knowledge/docs/en-US/Red_Hat_Enterprise_Linux/6/html/Virtualization_Host_Configuration_and_Guest_Installation_Guide/index.html Security Guide: https://access.redhat.com/knowledge/docs/en-US/Red_Hat_Enterprise_Linux/6/html/Virtualization_Security_Guide/index.html V2V Guide (import VMs from other vendors into KVM) https://access.redhat.com/knowledge/docs/en-US/Red_Hat_Enterprise_Linux/6/html/V2V_Guide/index.html You basically need to install libvirt and virt-manager on your box, and you're ready to go. To get SPICE support, you need to install some extra spice packages as well, which will improve the graphical console performance considerably (compared to VNC). All packages and dependencies are available in SL6. Give it a shot, and you'll see it's not necessarily that much harder than vbox. You can also run Windows guests within KVM. And there's even virtio drivers available [1] to improve the performance of disk and network IO. kind regards, David Sommerseth [1] http://www.linux-kvm.org/page/WindowsGuestDrivers/Download_Drivers On 10/2/12 3:15 AM, David Sommerseth wrote: - Original Message - From: Joseph Areeda newsre...@areeda.com To: SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV Sent: Tuesday, 2 October, 2012 12:33:59 AM Subject: The opposite SL and VirtualBox problem I want to run Windows as a guest system on my Sl6.3 box. Installing vbox from the Oracle repository gives me an error trying to create the kernel modules. Just a silly question. Why bother with VirtualBox when you have KVM built into the OS? Using the SPICE protocol (yum search spice) and you'll even get a decent console performance. And it's really easy to setup and configure using virt-manager. kind regards, David Sommerseth xxx
Re: [SCIENTIFIC-LINUX-USERS] Password required in single-user mode? (solved)
On 08/23/2012 03:18 PM, Pat Riehecky wrote: On 08/18/2012 03:57 PM, David Sommerseth wrote: Hi, I've been running Scientific Linux since the 6.0 days, and single-user mode have basically behaved how I have expected it those few times I needed it. As I usually set up my boxes root accounts with passwords disabled, single-user mode needs to be without root password. Today, after having upgraded from 6.3, I needed to enter single-user mode at boot. And I was asked for a password at boot time. Is this change intentional? # cat /etc/redhat-release Scientific Linux release 6.3 (Carbon) # rpm -qa | grep -i sl_password_for_singleuser | wc -l 0 # grep SINGLE /etc/sysconfig/init SINGLE=/sbin/sushell If this change was intentional, how can I go back to the old behaviour? I double checked the behaviour with an old VM with SL6.1, and that behaves as expected. kind regards, David Sommerseth Hi David, The behavior shouldn't have changed. You've provided just about all the relevant details in your email, so there isn't really anything I want to ask for more information. Can I have you try setting /etc/sysconfig/init = SINGLE to /sbin/sulogin rebooting and setting it back to /sbin/sushell? Perhaps something got 'stuck' wrong /sbin/sushell is a shell script, so can I have you verify its contents? Mine looks like: #!/bin/bash [ -z $SUSHELL ] SUSHELL=/bin/bash exec $SUSHELL Hi Pat, First of all, tanks a lot for your answer, and to Stephan for his input as well. I checked /sbin/sushell and the SINGLE variable in /etc/sysconfig/init. And it was currently working just as expected - in both modes. But as I knew it had been failing, I added an extra disk to the VM I have with SL6.3. Double checked adding 's' to the kernel command line gave me a shell without asking for password. Then I added this disk to /etc/fstab ... and removed the extra disk. This time, I was asked for a root password when booting up - both when I was asking for 's'ingle user-mode and without it. Adding the disk back, and the proper behaviour is back too. So that's the issue I hit, as I in my real setup had just added and configured a new disk with encryption on my VM - but messed up the encryption key. When I then wanted to rescue the system, instead of solving it via 'single-user mode' I had to use guestfish on the root filesystem of the failing machine and disable the new disk in /etc/cryptotab and /etc/fstab. Then I could start setting up the disk again with a proper key. But the hint from Stephan, setting EMERGENCY=/sbin/sushell in /etc/sysconfig/init did the trick. So that's what I need to set on all my boxes. Lesson learnt: Emergency mode supersedes single-user mode, especially with filesystem failures. Thanks all! kind regards, David Sommerseth
Re: SL 6 vs. other RHEL clones: security advisory comparison
On 20/08/12 14:02, Janne Snabb wrote: Hello, I made some statistics and comparisons about security advisories published by three popular RHEL 6 clones: CentOS 6, Oracle Linux 6 and Scientific Linux 6. The article is available at the following URL: http://bitrate.epipe.com/rhel-vs-centos-scientific-oracle-linux-6_187 I hope you find it interesting. This is really interesting. However, there are five Important and Critical erratas which are delivered considerably slower (RHSA-2012:0387, RHSA-2012:0388, RHSA-2012:1009, RHSA-2012:1054 and RHSA-2012:1064). SL's average is very much impacted by 3 of them. What would be more interesting, from a statistical point of view is to take those extremes out of the comparison. On an average level, All distros delivers errata updates fairly fast, but those extremes skews the real average. Of course extreme delays happened and will also happen in the future, I'm not trying to hide that. But getting an average what is more the typically expected average would be probably give a better indication. So my suggestion is to take those four erratas I listed out of the equation, which are only tagged as Important or Critical. Those five erratas are really a minority in a bigger set of data and surely looks like exceptions, across all distros. As I said, I'm not trying to hide the fact that SL had some slow deliveries. We need those graphs too. Such things happens to all distroes, sometimes you just get set behind. But they do actually skew the /typically expected/ delivery delay badly. And people with statistics background can probably explain even better than me why it's interesting to remove the extremes and compare that too. kind regards, David Sommerseth
Password required in single-user mode?
Hi, I've been running Scientific Linux since the 6.0 days, and single-user mode have basically behaved how I have expected it those few times I needed it. As I usually set up my boxes root accounts with passwords disabled, single-user mode needs to be without root password. Today, after having upgraded from 6.3, I needed to enter single-user mode at boot. And I was asked for a password at boot time. Is this change intentional? # cat /etc/redhat-release Scientific Linux release 6.3 (Carbon) # rpm -qa | grep -i sl_password_for_singleuser | wc -l 0 # grep SINGLE /etc/sysconfig/init SINGLE=/sbin/sushell If this change was intentional, how can I go back to the old behaviour? I double checked the behaviour with an old VM with SL6.1, and that behaves as expected. kind regards, David Sommerseth
Re: SL to 6.3 release
On 19/07/12 16:05, Federico Alves wrote: What I would like to see is a faster release of updates, somewhat close to the upstream. Uhm ... What about to make a guess ... Guess why SL/CentOS are without any costs and why RH increased their pricing For my part, I'd probably guess it's somewhat related ... I might be wrong, of course... kind regards, David Sommerseth
Re: is the drop to fsck visual fixed in 6.3?
On 18/07/12 23:02, Orion Poplawski wrote: On 07/18/2012 02:50 PM, Todd And Margo Chester wrote: I don't know the exact number, I think it is 27 reboots, your boot will automatically drop to an FSCK. In RHEL5, your would see a status bar showing you progress. In 6, you get no indication that an FSCK is happening and you think you are frozen. The temptation to throw the power switch is overwhelming. I can't speak to the lack of status, but you can disable the automatic fsck with: tune2fs -c 0 -i 0 /dev/ This is pretty much done automatically now for any filesystems that the install makes, but not ones that you create later. If using ext{2,3,4}, I would strongly recommend *against* doing this. Running fsck from time to time isn't a bad thing. It can surely happen that it finds some things which should be fixed. If this is needed fro xfs, I dunno. For reiserfs it is not needed, it will take care of this on its own - and really needed, it'll scream loudly and you won't be able to mount that partition/lv and this fix might take hours to complete. Regarding other file systems, I have no experience. Removing these fsck's is like skipping taking your car to a car check every now and then. You car may run for a long way without a service. But when it is really needed, it can take a long time to fix and might be more costly compared to if you did this regularly ... and if you want a reliable server, having a little maintenance downtime couple of times during a year might not be such a bad investment - in the long run. Just my 2cents kind regards, David Sommerseth
Re: is the drop to fsck visual fixed in 6.3?
On 19/07/12 20:20, Konstantin Olchanski wrote: I think fsck should be able to run in the background all the time. I am sure the ext2/3/4 filesystem driver and fsck can be made to talk to each other and permit checking a mounted live filesystem. (SUN ZFS can do this). This is actually partly possible if you use LVM. You create a snapshot of the file system, and run fsck on the snapshot. If no issues are found, no worries. If there are issues there, then you need to unmount the file system and run the fsck on the main image (not the snapshot image). And when that passes, you can delete the snapshot - this way you also have a backup in case something goes really bad. Anyhow ... this can save you downtime, when the fsck is clean, even an automatic fsck shouldn't cause much extra delay next time you boot. kind regards, David Sommerseth
Re: significance of indicating root partition location in bot process
On 18/07/12 16:28, anuraag chowdhry wrote: thanks for replying. but the files needed to boot i.e stage1, stage2, initram disk image and kernel are present in boot partition, right? and /boot can be on different hard disk partition too, right? stage1, stage2, the kernel and initramfs images are actually bootloader (GRUB, syslinux, lilo, etc) related. That's the needed pieces to put the kernel into RAM, boot it and enable the hardware needed to access harddrives and such. GRUB have also limited file system support, to be able to locate and load those mentioned files. When the kernel has started, the hardware is enabled, then it goes ahead and mounts the root file system (rootfs) which it takes from the root= kernel command line. When rootfs is ready, it starts the init process (upstart, /sbin/init or whatever the init= kernel command line argument says). This is the first process which should make sure to start the rest of the system processes ... like starting all the enabled services etc. It also takes care of enabling all the physical consoles (usually tty{1-6}) and respawning these consoles when you're logging out again. It's a little detail I haven't mentioned, but the very first step is to run some scripts which are inside the initramfs ... but that's a more advanced topic. are you referring to /sbin/init or /etc/syconfig/init files that reside in / partition ? if i remove the root= ,.. parameter, the system will try to boot to some extent , right? Nope, the kernel will panic, as it doesn't know about any rootfs. The rootfs usually needs at minimum /bin, /sbin and /etc. But it's possible to boot a system if you just have /etc available - but then the first init process (pid 1) needs to be in /etc as well. So to sum up the boot process: - You press the POWER ON button - Bootstrap/BIOS Selfcheck - BIOS POST starts - BIOS locates boot harddrive and loads the MBR boot loader (512 bytes) - MBR boots GRUB (stage1/stage2) - GRUB loads initramfs and kernel into RAM - GRUB boots the kernel - kernel executes scripts in the initramfs, prepares for mounting rootfs - kernel mounts rootfs and starts init process - init process mounts the rest of the file system(s) - init process continues loading/starting system config - init finally starts the console processes (Xorg, *getty) - The system is ready for logins (This is the typical Intel arch PC BIOS boot procedure, with EFI this changes a little bit. And non-Intel is quite more different) kind regards, David Sommerseth
Re: server crashing out of memory
% qemu-kvm 454820.14s 0.27s 0K 0K 0K 0K --- S 12 4% qemu-kvm 429990.05s 0.03s 0K 0K 0K 16K --- S 14 1% qemu-kvm 2092150.01s 0.05s 0K 0K 0K 0K --- S 9 1% qemu-kvm 404720.03s 0.02s 0K 0K 0K 0K --- S 12 1% qemu-kvm slabtop: Active / Total Slabs (% used) : 1599178 / 1599216 (100.0%) Active / Total Caches (% used) : 132 / 204 (64.7%) Active / Total Size (% used) : 6195130.43K / 6274921.86K (98.7%) Minimum / Average / Maximum Object : 0.02K / 0.28K / 4096.00K OBJS ACTIVE USE OBJ SIZE SLABS OBJ/SLAB CACHE SIZE NAME 7227321 6709793 92%0.10K 195333 37781332K buffer_head 4242650 4242336 99%0.07K 80050 53320200K selinux_inode_security 4224700 4224638 99%1.00K 10561754 4224700K ext4_inode_cache This smells a bit bad ... ext4_inode_cache is using a lot of memory ... 3257480 3257186 99%0.19K 162874 20651496K dentry 1324786 1250981 94%0.06K 22454 59 89816K size-64 484128 484094 99%0.02K 3362 144 13448K avtab_node 347088 342539 98%0.03K 3099 112 12396K size-32 342580 324110 94%0.55K 489407195760K radix_tree_node 236059 235736 99%0.06K 4001 59 16004K ksm_rmap_item 123980 123566 99%0.19K 6199 20 24796K size-192 105630 47803 45%0.12K 3521 30 14084K size-128 24300 24261 99%0.14K900 27 3600K sysfs_dir_cache 17402 15599 89%0.05K226 77 904K anon_vma_chain 16055 14874 92%0.20K845 19 3380K vm_area_struct 9844 8471 86%0.04K107 92 428K anon_vma 8952 8775 98%0.58K 14926 5968K inode_cache 7518 5829 77%0.62K 12536 5012K proc_inode_cache 6840 4692 68%0.19K342 20 1368K filp 5888 5532 93%0.04K 64 92 256K dm_io top - 10:10:02 up 22:34, 4 users, load average: 1.02, 1.15, 1.53 Tasks: 888 total, 1 running, 887 sleeping, 0 stopped, 0 zombie Cpu(s): 0.8%us, 1.2%sy, 0.0%ni, 97.9%id, 0.1%wa, 0.0%hi, 0.0%si, 0.0%st Mem: 49421492k total, 43619512k used, 5801980k free, 4409144k buffers Swap: 8388600k total,16308k used, 8372292k free, 25837164k cached Somehow, this doesn't reflect what the kernel complains about when the OOM killer starts its mission. I see that you're using kernel-2.6.32-279.1.1.el6.x86_64 ... that smells a bit like a SL 6.3 Beta ... is that right? As SL 6.2 is usually around 2.6.32-220-something. I would probably recommend you to try a 6.2 kernel if you're running something much more bleeding edge. And it somehow seems to be related to some file system issues ... at least from what I can see. Could be a bugy kernel which leaks memory, somewhere in either the parition table code or ext4 code paths. Not sure I'm able to provide any better clues right now. kind regards, David Sommerseth
Re: SL to 6.3 release
On 18/07/12 15:37, Semi wrote: 1) When the SL6.3 is expected? Centos 6.3 already exists. *sigh* do you read these mails in the thread you just replied to? Again, here's a couple of pointers which might give you a better clue: http://listserv.fnal.gov/scripts/wa.exe?A2=ind1207L=scientific-linux-usersT=0P=11071 http://listserv.fnal.gov/scripts/wa.exe?A2=ind1201L=scientific-linux-usersT=0P=17067 2) When I run update from Centos I also updating the release number. For example: from Centos 5.2 yum update I'll get Centos 5.8 and SL doesn't raises the release version ? That's documented here: SL 6.x: http://www.scientificlinux.org/documentation/howto/upgrade.6x SL 5.x: http://www.scientificlinux.org/documentation/howto/upgrade.5x Unless you have installed the yum-conf-sl6x package (on SL6.x), you won't upgrade the minor version of the distro automatically. So in that case, you need to do it explicit as described above. kind regards, David Sommerseth
Re: kernel-2.6.18-308.8.2.el5.x86_64 stalled at stage2 during boot
On 07/07/2012 02:09 AM, Zhi-Wei Lu wrote: Hi all, I have a Supermicro box (motherboard X7DW3) with 3ware RAID card (9690SA-4I). I have 24 drives attached to this raid card, while the first three drives were exported as SINGLE (JBOD) drives. The /boot and /root are Linux software raid 1 on these first three drives. The latest kernel-2.6.18-308.8.2.el5.x86_64 would failed to boot, but the previous kernel, kernel-2.6.18-308.8.1.el5.x86_64, worked just fine. I also tested with CentOS kernel-2.6.18-308.8.2.el5.x86_64 and ended up with the same fate. Does anyone see this problem at all? Maybe take a screen shot (with a camera) of the boot process where it stops, could be an idea? Make sure to remove 'rhgb' and 'quiet' from the kernel command line (you can edit the grub settings on the fly, if you reach for the menu). kind regards, David Sommerseth
Re: KVM issues with dump
On 07/06/2012 02:23 AM, Nico Kadel-Garcia wrote: By the way, if you skip using AHCI and just program the bios for IDE, you don't even have to use an f6 disk to install XP. No sign of drivers doing XP in yet. Although the day will come. Great letter. Thank you! You didn't notice any performance issues with virtualized IDE versus SCSI? IDE and SCSI emulation is not going to impress you on performance. Using virtio for disk access may improve the performance, and the Windows drivers should be available here: http://www.linux-kvm.org/page/WindowsGuestDrivers/Download_Drivers kind regards, David Sommerseth
Re: single user mode
On 06/23/2012 10:38 AM, anuraag chowdhry wrote: Hi I have SL6.1 and the moment , the second stage Grub presents the Kernel selection Menu, Pressing A ( for appending) and then a spacebar and then typing single , pressing ENTER , ultimately prompts me the root password after booting. is the behaviour not different here.? The system should boot in single user mode , it does but then it asks for root password for maintenace.or press Ctrl-D to continue. i tried typing '1 or s instead of single , but still it prompts for the password. my kernel is 2.6.32-131.0.15.el6.i686 Try booting with 'init=/bin/sh' in the kernel command line. This approach might force you to manually mount the root filesystem containing /etc ... and then you can change the root password, unmount and reboot. Then you can enter single user with a known password. If you save the root password stored in /etc/shadow before you change the password, you can also restore the old root password if that's needed afterwards. kind regards, David Sommerseth