[00:04] <adam_g> zul, jamespage https://code.launchpad.net/~openstack-ubuntu-testing/+activereviews <- in case you didnt see, ive proposed the 2013.1.3 updates there, since the grizzly branches are still hosted there. if you think we should migrate to ~ubuntu-server-dev, let me know and ill adjust accordingly
[00:11] <zul> adam_g:  yeah can we migrate them over
[00:23] <failmaster> it doesn't necessarily means that we have the device mounted on /boot during initrd stage? http://www.gnu.org/software/grub/manual/grub.html#prefix
[00:36] <twb> I'm after backported kernels for lucid.  I don't see them in lucid-backports.  Am I looking in the wrong place?
[00:41] <qman__> twb, there are backported kernels from maverick, natty, and oneiric in the form of linux-image-generic-lts-backport-oneiric
[00:41] <qman__> not sure if anything newer exists outside of PPAs
[00:42] <twb> Hm, I was looking in lucid-backports, not in PPAs.
[00:42] <qman__> those ones are actually in lucid-updates
[00:42] <twb> Ah.
[00:43] <twb> Yes, I see them now, thanks.
[00:45] <twb> That has 3.2, which I think will suffice.  If it's not, can you tell me where the ubuntuized git branches live?
[00:46] <twb> Aha, kernel.ubuntu.com looks promising
[00:50] <myhrlin> hi, I'm trying to open a program to X via ssh.  It used to be on certain systems I could simply DISPLAY=:0.0; firefox; or something
[00:50] <myhrlin> but now it's not working; would someone know how to accomplish this?
[00:51] <sarnold> myhrlin: try first, ssh -X hostname firefox   then ssh -X -Y hostname firefox
[00:51] <sarnold> the -Y enables less-secure options that might be required all the same
[00:51] <twb> Use -X, not -Y.  The manpage is misleading, -Y is for when you trust the network
[00:51] <sarnold> twb: oh?
[00:51] <myhrlin> sarnold: oh I'm not trying to forward an X session through ssh
[00:51] <sarnold> myhrlin: oh!
[00:52] <twb> sarnold: as you say, it's the less-secure version
[00:52] <twb> myhrlin: then 1) you really should; but if you insist; 2) learn about xhost and xauth.
[00:52] <sarnold> twb: I didn't think it had anything to do with network trust, but how much you trusted the X client on the remote machine
[00:52] <twb> sarnold: sorry, I am misremembering, you are right.
[00:53] <sarnold> twb: I believe myhrlin just wishes to display the contents on the remote machine as well...
[00:53] <twb> sarnold: the key point is a lot of people see "-Y trust" and thing it is safer than -X, which is backwards.
[00:53] <sarnold> twb: ah :) that's well worth correcting. :) hehe
[00:53] <sarnold> myhrlin: I'm pretty sure I've used DISPLAY=:0.0 foo    myself..
[00:53] <sarnold> myhrlin: try export DISPLAY  ?
[00:54] <myhrlin> sarnold: right, it does not seem to be working -- oh right... export
[00:54] <twb> myhrlin: if you want an X app to connect to a remote X server, you do "DISPLAY=example.net:0 xlogo"
[00:55] <twb> This will not work by default because example.net has to be told to allow connections from the app server, with xhost(1) and (or?) you have to export the MIT cookie from the X server and put it on the app server, using xauth(1).
[00:56] <twb> Unless you ultimately trust every single device on your network, you should use ssh -X instead, because otherwise the traffic is unencrypted and Bad Things can happen.
[00:57] <sarnold> twb: think "oh crap my window manager on my usual desktop is broken and I can't get a new terminal to run "killall windowmanager ; nohup windowmanager &"  -- I've restarted busted window managers via ssh before, bloody convenient thing to do once in a while. :)
[00:58] <sarnold> twb: I think myhrlin is attempting to solve a similar problem -- start a program on an existing X server from the same machine as the X server, but you're not sitting right at the keyboard that very second..
[00:58] <twb> sarnold: ah, then I misunderstood what he's trying to do.
[00:58] <sarnold> hehe
[00:58] <sarnold> twb: I may have misunderstood too, but :0.0 gives me hope I'm on the right track. :)
[00:58] <twb> sarnold: in your use case, you simply need to ensure DISPLAY is set and XAUTHORITY points to the cookie fiel
[00:59] <sarnold> myhrlin: ^^ note the bit about XAUTHORITY :) that might be another variable to try
[00:59] <twb> The latter can be skipped if it's in the default place (~/.Xauthority), which IIRC it is unless /home is on a network filesystem with root-squash
[00:59] <twb> In any case, myhrlin should get a meaningful error message to help us diagnose the problem further.
[01:37] <LargePrime> sarnold: no.  fighting with the openvpn guys
[01:37] <LargePrime> they not nice like you guys
[01:41] <LargePrime> he complaining cause i use netstat and not ss
[01:41] <LargePrime> but ss no work
[01:42] <twb> Then you're using it wrong.
[01:43] <LargePrime> ?
[01:44] <LargePrime>  netstat -an | grep 1194 works , ss -an | grep 1194 does nothing
[01:45] <LargePrime> what am i doing wrong?
[01:46] <twb> ISTR there is some obscure case where it doesn't list UDP properly
[01:46] <twb> I confess that's why I still use netstat.  That and it inserting ridiculous amounts of whitespace.
[01:46] <LargePrime> dam whitespace insertions
[01:47] <twb> FWIW I can reproduce the issue, can't remember the fix
[01:49] <twb> http://paste.debian.net/24509/
[03:00] <LargePrime> yay, it works
[03:17] <failmaster> could someone please help to shed some light on why this configuration stopped working with freshly installed 13.04 http://askubuntu.com/questions/330660/what-is-the-correct-way-to-use-key-file-luks-authorization especially curious if we have /boot accessible at it's absolute path on corresponding to cryptsetup boot stage?
[06:21] <lifeless> hallyn: ls ../20130812/var/lib/lxc/sid/rootfs/
[06:21] <lifeless> $
[06:22] <lifeless> hallyn: ... --one-file-system with rsync doesn't back up subvolumes that aren't mounted ;)
[07:31] <ejv> Need help installing mysql on 12.04 LTS : http://dpaste.com/1341424/
[07:32] <failmaster> ejv, purging it totally doesn't fix it?
[07:33] <ejv> man im trying
[07:34] <ejv> this is LTS, this shouldn't even be happening lol
[07:34] <ejv> when i purge it, it then tries to configure and install it, it's totally messed up
[07:35] <failmaster> ejv, so when you just `apt-get purge mysql*` it doesn't behave as it sounds?
[07:35] <ejv> right
[07:36] <ejv> i can't even dist-upgrade either
[07:36] <failmaster> i doubt that problem is about mysql =)
[07:36] <ejv> http://dpaste.com/1341426/
[07:36] <ejv> whatever updates are announced, I install, I usually don't review them
[07:36] <failmaster> vps?
[07:36] <ejv> dedicated
[07:37] <failmaster> ejv, maybe it would be a good idea to have a look at dmesg?
[07:37] <failmaster> because until you did everything keeping an eye on reccomendations like e.g. ones related to upgrade
[07:38] <ejv> tail
[07:38] <failmaster> mm*
[07:38] <ejv> tail:
[07:38] <ejv> http://dpaste.com/1341427/
[07:39] <ejv> when the purge happened, it knocked out the apparmor mysql profile
[07:39] <failmaster> ejv, no problems related to filesystem in there?
[07:39] <ejv> those init issues look... odd
[07:39] <failmaster> or apt cache or something
[07:39] <ejv> im not familiar with app cache
[07:39] <ejv> s/app/apt/
[07:39] <failmaster> syslog would be closer
[07:40] <ejv> i disabled apparmor
[07:40] <ejv> no idea what it might be doing
[07:41] <failmaster> before doing that?
[07:41] <ejv> syslog tail looks quiet
[07:42] <failmaster> look at the whole log files that would catch this accident timeline and never play with apparmor until you know what you're doing
[07:42] <ejv> im not playing with it i just 'service apparmor stop' hehe
[07:43] <ejv> ok i'll renable and backtrack, brb
[07:43] <failmaster> at least did you read the package description? lol
[07:49] <ejv> failmaster: i noticed my /boot is full
[07:49] <ejv> interesting...
[07:50] <failmaster> ejv, it happens due to the fact old stuff wasn't removed
[07:50] <ejv> purging kernels stand by :)
[07:52] <ejv> incredible, that was holding my entire system hostage
[07:54] <ejv> i don't fully understand how having space in /boot is a dependency for upgrading mysql-server
[07:56] <failmaster> this question could probably take place only after you'll make sure it was the reason
[07:57] <ejv> i cleared out some kernels, the panic went away
[07:57]  * ejv shrugs
[08:08] <ejv> thanks failmaster, all set now, good for another year i hope
[08:08] <failmaster> np you fixed it yourself! =)
[08:08] <ejv> it's my job during the day, at the end of the day, im just drained, this is the last thing i want to do ha
[08:09] <failmaster> well i'm lucky enough to not have a life at all lol
[08:10] <ejv> ;)
[08:51] <ivoks> when is the next server team meeting?
[09:12] <rbasak> ivoks: probably today at 1600 UTC.
[09:17] <ivoks> rbasak: thanks
[09:43] <ThothCastel> what does sudo apt-get install build-essential linux-headers-$(uname -r)   do?
[12:15] <hallyn_> lifeless_: I still don't understand what "arent mounted" means.  You don't manually mount a btrfs subvolume, it's just always there.  If you do "ls /var/lib/lxc/sid/rootfs", do you see the rootfs?
[12:16] <hallyn_> If not, then something doesn't match my understanding.  If so, then I'd say indeed this is an rsync bug (high prio, bc I expect it to affect lots of ppl)
[12:16] <hallyn_> (also, --one-file-system should ahve nothign to do with whether a sub-fs is mounted, another reason I'm confused :)
[12:23] <stemid> I have two identical 12.04 installs where one cannot run apt-get update. it gets 404 errors on many files, including http://se.archive.ubuntu.com/ubuntu/dists/precise-backports/universe/source/Sources and http://se.archive.ubuntu.com/ubuntu/dists/precise-backports/main/binary-amd64/Packages. in all the 404 cases the file eixsts with a .bz2 extension.
[12:23] <stemid> I've been trying to see what could differ on one host from the other, in the /etc/apt dir. but the files are identical.
[12:25] <stemid> not even if I generate a new list on http://repogen.simplylinux.ch/generate.php does it work. it must be missing some configuration.
[12:28] <stemid> http://www.fpaste.org/31763/37639692/
[12:29] <stemid> only from that server
[12:33] <stemid> $ host se.archive.ubuntu.com
[12:33] <stemid> se.archive.ubuntu.com has address 10.40.2.21
[12:33] <stemid> funny dns error =)
[12:33] <stemid> yet it's using google's public dns in resolv.conf and files dns in nsswitch
[12:34] <jpds> stemid: /etc/hosts ?
[12:34] <hallyn_> lifeless_: yeah confirmed, lemme ask on ubuntu-kernel
[12:34] <RoyK> stemid: same resolv.conf?
[12:34] <RoyK> oh
[12:35] <RoyK> jpds: the host command does a dns lookup, and doesn't care about /etc/hosts
[12:38] <stemid> I see now, the morons who installed it have installed a bind on it
[12:38] <stemid> so it's using its own bind
[12:38] <stemid> jesus
[12:38] <jpds> Every server should be running it's own bind.
[12:38] <jpds> its*
[12:38] <stemid> no
[12:53] <RoyK> jpds: you should have 2-3 servers running bind and then have the other servers using those
[12:54] <hallyn_> lifeless_: yeah, so it *is* a different fs according to rsync.  if you don't want to do any snapshot clones, we could add a flag to tell lxc not to make rootfs a subvolume.  (failing that the only 'fix' apart from changing your backup script woudl be to teach rsync something - not sure what)
[12:54] <hallyn_> lifeless_: was that what you were after originally?  :)
[13:09] <jpds> RoyK: I was joking about bind everywhere.
[14:01] <jamespage> adam_g, zul: seen this before? - http://paste.ubuntu.com/5981288/
[14:01] <jamespage> grizzly cloud archive
[14:17] <zul> jamespage: https://code.launchpad.net/~zulcss/python-novaclient/2.14.1/+merge/179929
[14:17] <jamespage> zul: seen this before? - http://paste.ubuntu.com/5981288/
[14:18] <jamespage> grizzly cloud archive on precise
[14:18] <zul> jamespage:  no i havent
[14:19] <jamespage> zul, keystone catalog is OK
[14:19] <jamespage> but keystone endpoint-list borkes
[14:19] <jamespage> infact *-list borkes
[14:19] <zul> which versoin is this with?
[14:20] <jamespage> zul, whatever is in grizzly CA
[14:20] <jamespage> zul, might be some sort of client issue I think
[14:20] <jamespage> zul, as my keystone client on saucy works fine against this deployment
[14:21] <jamespage> but the keystoneclient within the deployment breaks
[14:21] <jamespage> which is breaking openstack-dashboard
[14:21] <jamespage> zul, curl'ing the commands works fine
[14:25] <zul> jamespage:  possible bug in webob?
[14:27] <jamespage> zul, maybe - if I comment out the offending debug code everything springs back to life
[14:27] <jamespage> odd
[14:28] <zul> jamespage:  odd indeed
[14:28] <jamespage> feels like a bug
[14:37] <jamespage> zul, spotted the problem - the environment was running essex keystone
[14:37] <jamespage> that sucks
[14:37] <jamespage> upgraded and it fixed the problem
[14:37] <zul> jamespage:  gah?
[14:37] <freze> hi all, is it safe to upgrade the kernel on ubuntu LTS?
[14:37] <jamespage> zul, I was helping marco with a demo for tomorrow
[14:38] <zul> jamespage:  ah ok
[14:40] <zul> jamespage:  https://code.launchpad.net/~zulcss/swift/1.9.1/+merge/179933
[14:40] <jamespage> zul, not going to get to that for a bit - working on a nasty-ish golang issue right now
[14:40] <zul> jamespage:  ok
[14:41] <jamespage> zul, oh - go on then
[14:41] <jamespage> won't take long I guess
[14:41] <zul> jamespage:  ill find a willing guinea pig
[14:42] <jamespage> https://code.launchpad.net/~zulcss/python-novaclient/2.14.1/+merge/179929
[14:42] <jamespage> zul, ^^ that is targetting the wrong branch btw
[14:43] <zul> jamespage:  fixed
[14:45] <zul> jamespage:  fixed swift
[14:54] <zul> roaksoax: https://code.launchpad.net/~zulcss/python-cinderclient/1.0.5/+merge/179938
[15:04] <roaksoax> zul: done
[15:04] <zul> roaksoax:  thanks
[15:11] <zul> roaksoax:  one more
[15:11] <zul> https://code.launchpad.net/~zulcss/python-ceilometerclient/1.0.3/+merge/179940
[15:12] <roaksoax> zul: done
[15:12] <zul> roaksoax:  thank
[15:57] <jamespage> hallyn_, can you run todays IRC meeting? I see smoser in the chair but he's out today
[15:57] <hallyn_> jamespage: i really can't...
[15:57] <jamespage> hallyn_, np - next!
[15:57] <jamespage> roaksoax, ^^
[15:58] <anomaly> I am using a laptop temporarily for testing.  to cover all the options, would the server installation detect this and use some sort of power management?  I have noticed at times I can not ssh in, but when I tap a key on the keyboard I can.
[15:58] <jamespage> arosales, around - you might be chair by default if so
[15:59] <arosales> jamespage, hello
[16:00] <arosales> smoser, hallyn, and roaksoax all out huh
[16:00] <rbasak> I think it might be zul's turn again as he was the last person not to do the minutes, AFAICT.
[16:00] <arosales> jamespage, I can chair :-)
[16:00] <jamespage> arosales, great
[16:00] <arosales> uh oh zul
[16:00] <jamespage> zul, did you not do the minutes?
[16:00] <zul> rbasak:  uh...no i hcnaged it
[16:01] <arosales> ah I'll take zul's word for it then :-)
[16:01] <rbasak> Yeah the rota was changed, but https://wiki.ubuntu.com/MeetingLogs/Server is way behind
[16:01]  * arosales will chair this go around
[16:01] <zul> im *not* chairing
[16:01] <arosales> I'll action zul to update minutes
[16:01] <jamespage> arosales, thanks
[16:01]  * arosales pulls up instructions, one sec
[16:02] <arosales> #startmeeting ubuntu-server-team
[16:02] <arosales> sorry wrong window
[16:03] <roaksoax> jamespage: sorry just saw your ping... internet connection crapped on me :)
[16:06] <jamespage> roaksoax, np - arosales stepped into the breach!
[16:54] <adam_g> jamespage, do you reckon we should move grizzly branches from ~openstack-ubuntu-testing to ~ubuntu-server-dev? or keep them where they are?
[17:01] <zul> hallyn_:  how do you get out of of the lxc console again
[17:36] <hallyn_> zul: ctrl-a q
[17:36] <zul> hallyn_:  ahhh
[17:39] <hallyn_> zul: it *does* say that in the lxc-console manpage fwiw
[17:40] <hallyn_> (so i dont *totally* suck :)
[17:40] <zul> hallyn_:  man pages *ppphppt* ;)
[17:41] <hallyn_> 'those are so 1980'
[17:57] <hallyn_> lifeless_: too early to scream victory, but i believe the source of the libvirt leak has been foudn - there was a leak in netcf!  i'm running for the afternoon with Eric Blake's patch to fix that
[19:32] <hallyn_> ahs3: hey, need to push a set of netcf packages (memleaks all-around).  just a heads-up
[19:47] <ahs3> hallyn_: groovy.  just say when...
[19:51] <hallyn_> jessie has thrown me for a loop.  well lemme start with wheezy
[19:54] <lifeless> hallyn_: cool about the netcf fix; I will test packages if you want
[19:55] <hallyn_> lifeless: thx, i was gonig to do the debianpackage first to make sure they can stay in sync, so it'll be a bit yet
[20:29] <digia> Quick 12.04 question. Server crashed last night and today when logging in i get a login loop. I am able to login via recovery and create a seperate account, which i am able to log into. Where would i look to fix the login loop for the main account?
[20:56] <Fuzzbrain> console or X11 looping?
[21:32] <mgriffin> anyone familiar with phpMyAdmin on 12.04? these docs seem wrong https://help.ubuntu.com/12.04/serverguide/phpmyadmin.html because i think i need to modify /etc/dbconfig-common/phpmyadmin.conf and then dpkg-reconfigure phpmyadmin
[21:35] <mgriffin> oh, i figured it out.
[21:36] <mgriffin> the things i place in /etc/dbconfig-common/phpmyadmin.conf and regenerate /etc/phpmyadmin/config-db.php with aren't parsed if i leave dbc_dbname empty in /etc/dbconfig-common/phpmyadmin.conf, this doesn't make any sense ;)
[21:38] <moldy> hi
[21:38] <moldy> updated a 10.04 system to 12.04. rootfs on lvm on dmraid. boot fails. i think that it's missing the lvm modules within the initramfs (looking at the initramfs image, i cannot find them). how should i fix this?
[21:43] <mgriffin> moldy: did you read any of the docs?
[21:43] <moldy> mgriffin: i tried googling for the problem, but i was not able to find a solution yet
[21:43] <mgriffin> you can boot off of a live cd and mount, yeah?
[21:44] <mgriffin> i have fixed similar on rhel before by booting a rescue, bind mounting all of the things and then chroot, finally update-initramfs
[21:44] <moldy> mgriffin: yep
[21:44] <mgriffin> (bind mount /boot /proc /dev etc inside where you will chroot)
[21:44] <mgriffin> perhaps that is overkill, not sure
[21:45] <moldy> i mounted /proc and /dev from a grml live pendrive
[21:45] <moldy>  /boot is not on a separata partition. i should not mount the live medium's /boot, should i?
[21:47] <mgriffin> so you have in your livecd /mnt/the_local_disk/etc/fstab
[21:47] <mgriffin> like the local system is mounted somewhere?
[21:47] <moldy> after booting from the live disk, i run mdadm -A --scan and vgchange -a y
[21:48] <moldy> after that, i mount my lv to /mnt/ubuntu
[21:48] <moldy> but i think my current problem is that i am missing the lvm modules
[21:49] <mgriffin> did you use any bind mount after mounting /mnt/ubuntu (i don't know if this is necessary)
[21:49] <moldy> proc and dev
[21:49] <mgriffin> so you did bindmount /dev /mnt/ubuntu/dev
[21:50] <moldy> i have "lvm2" installed, but i cannot find any lvm modules on the system
[21:50] <moldy> yep
[21:50] <mgriffin> k, did you look in your /mnt/ubuntu/usr/share/initramfs/hooks/
[21:51] <moldy> there is a file called "lvm2" in /usr/share/initramfs-tools/hooks
[21:51] <mgriffin> or i guss /usr/share/initramfs-tools/hooks/
[21:52] <moldy> lvm does use kernel modules, right
[21:52] <moldy> ?
[21:54] <moldy> ah, it's called "dm-mod"
[21:58] <moldy> i have all sorts of dm-* modules, but no dm-mod.ko
[22:05] <mgriffin> i think this is pretty clear to follow: https://help.ubuntu.com/community/UbuntuDesktopLVM
[22:06] <mgriffin> so maybe cat /etc/initramfs-tools/modules to see what you have
[22:14] <moldy> mgriffin: will double-check, thanks
[23:11] <hallyn_> ahs3: alas there's a problem.  On jessie i can't build netcf as is.  gnlib issue, gives me:
[23:11] <hallyn_> ./stdio.h:1012:1: error: 'gets' undeclared here (not in a function)
[23:11] <sarnold> gets(3)? o_O
[23:12] <sarnold> I thought we burned that with fire?
[23:13] <hallyn_> sarnold: <shrug>  i dunno what gnulib is doing
[23:14] <sarnold> gets.c:6:2: warning: ‘gets’ is deprecated (declared at /usr/include/stdio.h:638) [-Wdeprecated-declarations]
[23:15] <sarnold> guess we -didn't- burn it with fire. darn.
[23:15] <hallyn_> sarnold: well this is in jessie
[23:15] <hallyn_> i don't expect it's the case in saucy - though i've not gotten to that yet
[23:19] <ahs3> hallyn_: hrm.  i reckon it needs removing...gets() is definitely on the Very Bad Thing To Do list
[23:25] <hallyn_> ahs3: it's in /usr/include/stdio.h
[23:25] <hallyn_> uh no
[23:25] <hallyn_> but it's ingnulib.  so i can't really remove it afaik
[23:34] <ahs3> hallyn_: oh, sorry.  i was thinking it could be removed from netcf, not the lib.  nm.
[23:35]  * ahs3 got distracted by a phone call
[23:44]  * hallyn_ got distracted by a bowl bearing delicious hot ramen breaking :(