[01:36] <adam_g> win 17
[03:49] <stgraber> hallyn: would you consider adding shutdown.conf to lxcguest in the ppa for lucid?
[04:08] <Zac_o_O> hello!  The drives on my server spin up every time a windows PC on the network is woken up from sleep.  This PC backs up over SMB.  How can I make this not happen?
[04:13] <Zac_o_O> It's driving me crazy!
[04:16] <twb> not spinning them down would be one way
[04:17] <twb> Making sure the windows PC is speaking DNS to the DNS resolver, instead of speaking NetBIOS to the samba server, might help.
[04:17] <twb> It depends why it's talking to the server on resume
[04:20] <Zac_o_O> how would i know if it's speaking DNS?
[04:20] <twb> you go into its registry somewhere
[04:20] <twb> IIRC DNS is the default since about vista
[04:21] <twb> Prior to that it would prefer netbios
[04:21] <twb> ##windows probably can help with that
[04:21] <twb> And/or do packet sniffing
[04:21] <Zac_o_O> ah.  ok.  is there anything I can do on the server side?
[04:21] <twb> 14:21 <twb> And/or do packet sniffing
[04:23] <Zac_o_O> hm.  so no change in a smb config file eh?
[04:41] <uvirtbot> New bug: #973953 in nova (main) "euca-describe-instance returns error VolumeNotFound" [Undecided,New] https://launchpad.net/bugs/973953
[05:03] <koolhead17> morning all
[05:05] <Zac_o_O> morning?  late here :)
[05:09] <koolhead17> Zac_o_O: i am in India :P
[07:46] <lynxman> morning o/
=== sergevn is now known as zz_sergevn
=== zz_sergevn is now known as sergevn
[09:00] <jasonmsp> hey all.  I've installed munin and nagios on our server. Its been great.  But I would like to see which programs are contributing most to the CPU%.  Is there a way to do this with, or without either of these?
[09:02] <lynxman> jasonmsp: I reckon something like this might be useful http://serverfault.com/questions/185804/monitor-and-graph-cpu-usage-per-process-and-per-thread
[09:02] <lynxman> jasonmsp: unfortunately there's no such default plugin for munin, so as it says on the post, enabling process accounting then writing a very quick munin plugin would be the most convenient option
[09:04] <jasonmsp> lynxman: thanks
[09:05] <lynxman> jasonmsp: my pleasure :)
[09:10] <Oxygen02> hello everyone. Running ubuntu-server 12.04LTS beta. I've to add a repo to the sources.list deb http://repo.percona.com/apt lenny main. is it lenny or?...
[09:10] <Oxygen02> cat /etc/debian-version shows me something different.
[09:12] <Oxygen02> it shows. wheezy/sid  What shoud I add as repo? anyone?
[09:41] <__sjh> lo all, just tried in #ubuntu to no avail, its kinda noisy, figured i'd try here too:
[09:41] <__sjh> loha, using 10.04, trying to change locale have followed https://help.ubuntu.com/community/Locale, but after running update-locale lang=en_GB.UTF-8, relogged, then run locale, now i see LANG= .... and nothing else, any ideas? ... just noticed i give en_GB.UTF-8 to locale-gen, but locale -a reports en_GB.utf8, any significance to that?
[10:20] <__sjh> if anyone is listening, if i login as root I get the locale i want, but as a normal user i dont ... any ideas how that could happen following on from previous comment?
[10:49] <Vivek> nick58b: Are you around ?
[10:50] <Vivek> nick58b: Please do check your pm.
=== matsubara-afk is now known as matsubara
=== leonel2 is now known as leonel
[12:01] <uvirtbot> New bug: #974172 in lxc (universe) "lxc-ls show duplicated names" [Undecided,New] https://launchpad.net/bugs/974172
[13:10] <Rapid2214> Hi, does anyone have a HP NC375T PCI Express Quad Port Gigabit Server Adapter working in a bond > bridge ? My bond works with my internal NICs but a bond on this card fails - I have seen bug reports but cannot find drivers to replace it
[13:19] <patdk-wk> Rapid2214, should work fine, you have more details?
[13:21] <Rapid2214> Its using the netxen_nic - Im using bond > Bridge with VLANS - The bond I have works with my VMs if I am not using this NIC, thanks :)
[13:25] <Rapid2214> Onboard drivers are using the bnx2 drivers
[13:29] <Rapid2214> From mii-tool: SIOCGMIIPHY on 'eth5' failed: Operation not supported
[13:32] <hallyn> stgraber: sure (shutdown.conf in lucid lxcguest) - note lxcguest is purely in ppa in lucid, btw
[13:46] <hallyn> hm, my backgrounds have gone away and been replaced by way too bright ones.  So I'm back to a black background.  Old times!
[13:47] <hallyn> stgraber: ok i'll go ahead and push shutdown.conf to lxcguest in lucid today.
[13:48] <hallyn> it's definately not going into upstart there?
[13:48] <stgraber> hallyn: thanks
[13:48] <stgraber> hallyn: no, it's a new "feature" so won't ever get into an update
[13:48] <hallyn> is it?  i would consider it a bugfix :)
[13:49] <stgraber> hehe, I guess you could argue that ;)
[13:49] <stgraber> if you want to be safe, rename it to lxc-shutdown.conf to avoid potential conflicts
[13:49] <hallyn> good point
[14:02] <uvirtbot> New bug: #974256 in nova (main) "Exception during the installation of Nova Essex-rc4 on Ubuntu 12.04" [Undecided,New] https://launchpad.net/bugs/974256
[14:02] <jcastro> matsubara: any testing feedback from people? I have one
[14:02] <matsubara> jcastro, nope
[14:04] <hallyn> stgraber: pushed
[14:33] <jdstrand> SpamapS: hi! what are your plans with mysql-5.1 in precise? it is currently in universe and trails behind oneiric. I thought I remembered reading it was going to be removed
[14:41] <smoser> rbasak, around ?
[14:42] <rbasak> smoser: yes
[14:42] <smoser> you think you migh thave time to improve the fix for bug 971820 ?
[14:42] <uvirtbot> Launchpad bug 971820 in squid-deb-proxy "squid-deb-proxy needs special handling of Release, Packages, Source" [Undecided,In progress] https://launchpad.net/bugs/971820
[14:42] <smoser> i'm concerned that what we have there does not really fix the development release issue.
[14:42] <rbasak> I'd love to but I need to better understand how it's broken first
[14:43] <smoser> so you think that the change you put in makes squid check the source headers versus what it has cached, right?
[14:43] <smoser> what headers explicitly does it check ?
[14:44] <rbasak> I don't know - squid calls it a "refresh" and supposedly does the right thing
[14:44] <rbasak> I can investigate that
[14:45] <smoser> well, whatever it does, i had a case where the server had new stuff, but squid was caching old stuff.
[14:45] <smoser> (and the old stuff was bad)
[14:45] <rbasak> what were the timestamps?
[14:45] <smoser> one possibility is that the server is busted. that it does not correctly update those timestamp.s
[14:45] <smoser> yeah... it sucks.
[14:45] <smoser> i'm  abad bug report
[14:45] <smoser> i know
[14:46] <rbasak> As in had the server's new stuff really gone up in timestamp, and was the previous fetch with a previous timestamp
[14:46] <rbasak> I've got one thought about this
[14:46] <rbasak> I could run a cron job
[14:46] <smoser> i stopped trying to find the correct fix after losing 2 hours on it. and realizing that orchestra had done the deny business.
[14:46] <smoser> the one thought i had is to just hard code the development release to deny
[14:46] <smoser> but let others go to cache
[14:47] <rbasak> That hits the archive with squid and without squid (with current squid-deb-proxy) and lots the timestamps and md5s
[14:47] <smoser> oh, and always deny -updates and -security
[14:47] <rbasak> logs
[14:47] <smoser> right. to catch it.
[14:47] <smoser> that would be good.
[14:47] <rbasak> yep
[14:47] <rbasak> then we'd also hopefully have the timestamp information which should help pin it down
[14:48] <smoser> but the hack of "if devel release, or -updates or -security or -backports, then deny"
[14:48] <SpamapS> jdstrand: yes, it should be removed. I beliee the rdeps are all gone.
[14:48] <smoser> would cache the largest hunk of data (as the main repo is the biggest by far)
[14:48] <SpamapS> jdstrand: believe.. even
[14:48] <SpamapS> jdstrand: re mysql 5.1
[14:49] <rbasak> does the dev archive tend to move as fast over the weekend as it does during the day?
[14:49] <jdstrand> SpamapS: would you mind double checking that, and filing a bug for its removal?
[14:49] <smoser> rbasak, just make sure you store the headers that you get back in both requeswts. and i think you'll get enoug hinfo to catch it.
[14:49] <smoser> oh, and you might as well pin your server too
[14:49] <jdstrand> SpamapS: feel free to assign it to me
[14:49] <smoser> ie, avoid the round-robin failures.
[14:50] <SpamapS> jdstrand: looks like libmyodbc and sphinxsearch are the last hold-outs.. will investigate today
[14:50] <jdstrand> SpamapS: awesome, thanks!
[14:52] <rbasak> smoser: I'm going to leave archive.u.c unpinned. I'll test ports.u.c in parallel which is effectively pinned because there is only one. That gives us both because I want to understand how well it works unpinned as well. I will only be donwloading the files - not giving them to apt - so the Release/Packages mismatch won't matter.
[14:53] <rbasak> smoser: do you think I need to download the whole file for the non-cached version, or will a direct HEAD request do? I think we can trust the E-Tag, can't we?
[14:53] <smoser> but ports.u.c gets updated less frequently i suspect.
[14:53] <smoser> ping the host
[14:53] <smoser> you'll get random failures we're not trying to address if you do not.
[14:53] <smoser> rbasak, i dont know. i think yo might need to download the file
[14:53] <smoser> to verify that it is different.
[14:54] <smoser> i guess the headers shoudl tell you that.
[14:54] <smoser> but the only way i can explain the issue is bad headers.
[14:57] <rbasak> smoser: what makes you think it was a squid issue rather than a mirror out of sync with itself?
[14:57] <smoser> because if i used check-archive at the same time, it was fine
[14:58] <smoser> and, actually... i tried to get check-archive to fail going through squid, but could not.
[14:58] <smoser> the reason for that is the way that check-archive works. it resolves the IP address and sets 'Host: ' header.
[14:58] <smoser> so squid didn't then recognize that 'http://archive.u.c' was the same as what check-archive was asking for.
[14:59] <smoser> and check-archive got a refreshed versions.
[15:00] <smoser> i debugged that a bit, but in squids logs, check-archive's hits were logged by server IP address, where the 'apt-get update' was getting dns name.
[15:02] <jcastro> robbiew: best MAAS comment so far: http://www.reddit.com/r/Ubuntu/comments/rufub/metal_as_a_service_canonical_announces_ubuntu/
[15:04] <chmac> I'm seeing incredibly slow disk performance on one of our servers hosted with OVH.
[15:04] <robbiew> jcastro: heh
[15:04] <chmac> I'm not sure how to investigate. Looking at the output of `sudo iostat -k 15 12` shows disk writes about 500K/s.
[15:05] <chmac> Any suggestions on how I can debug what's causing the slow disk access?
[15:05] <chmac> cpu utilisation looks like 25% iowate, 75% idle
[15:09] <Rapid2214> chmac: Is it a VM? or dedicated?
[15:09] <chmac> Rapid2214: Dedicated hardware, two SATA2 2T disks running software raid
[15:10] <Rapid2214> Have you checked for a degraded disk? /proc/mdstat ?
[15:11] <chmac> Rapid2214: Not sure how to read the output, nothing screams error to me...
[15:11] <chmac> I sent the smartctl output of each disk to the host, but waiting for a response, also not sure how to read that.
[15:12] <Rapid2214> chmac: [UU] means both disks are good
[15:12] <chmac> Rapid2214: Both disks are good
[15:12] <Rapid2214> anything in dmesg?
[15:14] <chmac> Rapid2214: Anything I could grep for? Not familiar with the output, so not sure what to look for...
[15:16] <chmac> Rapid2214: Thanks a lot for your help btw, I'm at a bit of a loss on this one.
[15:16] <Rapid2214> chmac: Would be obvious, if anything was wrong, constant messages is a tail tail sign, any processes in perticular which are using high wait? Is it a new server? Or has it been in use a while? No probs :)
[15:17] <chmac> Rapid2214: It's a reasonably new server, I think we've had the issue since it was brought online.
[15:17] <chmac> Rapid2214: dmesg looks good to me, `tail -f /var/log/dmesg` doesn't show anything new being added.
[15:18] <chmac> Rapid2214: The issue seems to be primarily with MySQL because it's the biggest user of disks.
[15:18] <chmac> Rapid2214: I had thought to try some kind of test copying data between partitions just to see what's going on.
[15:19] <Rapid2214> chmac: I have a MySQL server on software raid, personally i wouldn't recommend it, but 20% is terrible for that application tbf - What does # hdparm -tT /dev/<your-main-disk>  ----- return?
[15:20] <Rapid2214> *isn't terrible
[15:25] <chmac> Rapid2214: Ok, I think I'm onto a red herring, I just copied a 2G file from one partition to another and saw one report of 40M/s to md2
[15:25] <chmac> Seems like it's an issue with MySQL somehow then, that's what got me started on this, mysql is abysmally slow.
[15:26] <chmac> Although, a 2G file takes longer on this machine than on another single disk machine of a much lower spec (ram/cpu).
[15:27] <Rapid2214> chmac: Humm yeh, thats the issue I seem to have, its a backup server so it's not so important - good luck
[15:27] <chmac> Rapid2214: Same here, it's also a backup machine.
[15:27] <chmac> Rapid2214: You have the same issue where mysql is abysmally slow on software raid?
[15:29] <Rapid2214> chmac: Its noticeably slower with high waits, Im on software raid, but have put it down to old SCSI disks - I'm gonna be virtualising it soon
[15:29] <chmac> Rapid2214: It took me >1hr to import a 1.7M file that takes <2 minutes on our primary server, and 8m on a crappy £17 a month dedicated box.
[15:29] <chmac> Rapid2214: The mysql slave is 44h behind master :-(
[15:33] <Rapid2214> chmac: That is impressive, have you been using mk toolkit?
[15:34] <chmac> Rapid2214: I'd like to install mk_hearbeat, but wasn't aware of the toolkit
[15:38] <Rapid2214> chmac: Its quite good, it allows checksum checks, and sync to fix replication issues
[15:39] <Rapid2214> chmac: Its quite good, it allows checksum checks, and sync to fix replication issues
=== matsubara is now known as matsubara-lunch
=== hazmat is now known as kapilt
[15:59] <PhotoJim> normally Ubuntu does that shallow reboot that lets it do a kernel reload on server edition.  but how do you force a proper full reboot, if desired?
=== jtv1 is now known as jtv
[16:05] <mgw> for use in d-i partman, how can I determine the partition id (for a raid recipe)?
[16:06] <smoser> PhotoJim, i'd look in the kexec-tools man pages
[16:07] <mgw> e.g., in this example it uses sdX1, sdX5, etc
[16:07] <mgw> http://d-i.alioth.debian.org/manual/en.amd64/apbs04.html
[16:08] <smoser> PhotoJim, 'coldreboot'
[16:08] <smoser> holy crap.
[16:08] <smoser> and i just look at the implementation of that.
[16:09] <smoser> it looks like it can easily be changed by /etc/default/kexec, but as it is
[16:09] <smoser> NOKEXECFILE=/tmp/no-kexec-reboot
[16:10] <smoser> any user can touch that file, and next reboot will not be kexec.
[16:11] <PhotoJim> smoser: thanks much.
[16:11] <PhotoJim> k
[16:11] <smoser> PhotoJim, but the right way is to use 'coldreboot'
[16:11] <PhotoJim> smoser: greatly appreciated.
[16:31] <mgw> partman-auto-raid: mdadm: device /dev/sda5 not suitable for any style of array
[16:32] <mgw> any ideas what exactly that means? Here is my preseed:
[16:32] <mgw> https://gist.github.com/d3be3b4e2b5a9ec9221a
[16:33] <Rapid2214> mgw: Has the superblock been cleared?
[16:34] <mgw> How would I know?
[16:35] <mgw> I don't see anything about superblocks in rsyslog
[16:35] <Rapid2214> Oh, i don't know that program, but i would imagine it could be something like that
[16:37] <Rapid2214> mgw: Does the program run inside of a running ubuntu box?
[16:37] <mgw> no, it's a preseed
[16:37] <mgw> for the installer
[16:38] <mgw> there's no OS on the system at all
[16:38] <Rapid2214> Oh ok, do you get any other errors before that one?
[16:40] <hallyn> smb: sorry... to be clear, I should add https://launchpadlibrarian.net/99981918/no-vfb-for-hvm.debdiff  to my next libvirt push?
[16:40] <smb> hallyn, Yes, sorry for the confusion. We now agree to agree.
[16:43] <hallyn> ok
[16:47] <zul> hallyn:  404 on the libvirt-rbd debdiff
[16:53] <hallyn> hm
[16:54] <hallyn> zul: retry?
[16:55] <zul> hallyn: got it now
[16:55] <hallyn> zul: weird thing is it had a one-letter typo, but it was all done with cut/paste!
[16:55] <hallyn> perplexing
[16:55] <zul> hallyn: doesnt the ceph stuff need to be enabled in libvirt?
[16:55] <hallyn> <shrug>
[16:56] <hallyn> i don't get why they want it
[16:56] <hallyn> maybe they just want to have a ppa with it enabled but with exact same source as the archive version?
[16:56] <hallyn> or maybe it does make a difference
[16:56] <hallyn> they said it makes their testsuite pass
=== matsubara-lunch is now known as matsubara
[16:57] <zul> o...k
[16:57] <hallyn> ack
[16:58] <zul> well if it doesnt mess with anything sure why not
[16:58] <hallyn> right, it takes a few hours away from me but other than that is fine
[16:58] <hallyn> (i'm using the rig on which i was going to start the kvm perf tests, to run regression test)
[16:59] <hallyn> s/rig/wussie laptop/
[16:59] <zul> heh...im ok with it check with daviey
[16:59] <hallyn> yup, i did
[16:59] <hallyn> thx
[17:05] <zul> SpamapS: around?
[17:05] <zul> SpamapS: unping
[17:09] <SpamapS> zul: mooping
[17:09] <zul> SpamapS: i think i have a juju problem
[17:10] <zul> SpamapS: i have a ec2 keypair that i use for instances on openstack that doesnt seem to being used when i do juju status
[17:10] <esuave> is there a way i can lock the crontab while im editing, so nobody else can edit the same crontab and overwrite my changes?
[17:15] <SpamapS> zul: how did you tell juju about them?
[17:15] <SpamapS> zul: and, maybe -> #juju ?
[17:15] <zul> hold on
[17:15] <hallyn> esuave: i thought crontab in olden days used to flock...  i suppose you could wrap crontab in a script that flocks a file to make that happen
[17:16] <esuave> hallyn: yeah.. ill have to check it out
[17:16] <esuave> hallyn: running 8.04
[17:16] <hallyn> esuave: it doesn't do it in 12.04 either.  by olden days i was thinking sunos 4.x
[17:17] <esuave> ahh.. yeah thats a pisser.. cause while im editing.. someone else can edit the same crontab and overwrite my changes if they finish after me
[17:20] <stgraber> hallyn: I updated https://wiki.ubuntu.com/UDS-Q/Plenaries with details on the LXC plenary, I'll probably poke you to review the details when I start preparing it
[17:20] <hallyn> lxc plenery?
[17:21] <uvirtbot> New bug: #974450 in irqbalance (main) "irqbalance classifies network interfaces with custom/renamed interfaces as class other" [Undecided,New] https://launchpad.net/bugs/974450
[17:21] <hallyn> stgraber: cool, thanks :)
[17:33] <esuave> ok so im trying to create an alias.. is there a way i can do an IF this file exists.. execute this command?
[17:34] <esuave> like alias='if /home/test.txt exists; execute crontab -e'
[17:34] <esuave> i just dont know they syntax lol
[17:34] <esuave> the* not they
[17:36] <MarKsaitis> so can somebody help me? Or nobody knows a solution? I have vnc4server (tried tightvncserver too) installed, using hidden virtual desktops (the default). When I click logoff in my session, IT DOESNT DO A DAMN THING, DOESNT LOGOFF! FTW? Ubuntu 11.04 LXDE!!!
[17:39] <MarKsaitis> I bet nobody knows, as usual
[17:39] <patdk-wk> hmm, this is the server channel, we don't deal in gui
=== kapilt is now known as hazmat
[17:41] <hallyn> i wonder...
[17:41] <hallyn> !gui
[17:41] <ubottu> The graphical user interface (GUI) in Ubuntu is composed of many elements, including the !X server, a window manager, and a desktop environment such as !GNOME or !KDE (which themselves use the !GTK and !Qt toolkits respectively)
[17:42] <hallyn> not *particularly* helpful
[17:42] <hallyn> looks like #lxde is on irc.oftc.net
[17:51] <zul> utlemming: ping http://paste.ubuntu.com/916324/ whats nov.ec2?
[17:53] <utlemming> zul: I have no idea...what's your cloud-config look like?
[17:54] <zul> /etc/cloud/cloud.cfg?
[17:54] <utlemming> zul: /var/lib/instance/cloud-config.txt
[17:54] <utlemming> er.../var/lib/cloud/instance/cloud-config.txt
[17:56] <zul> utlemming: http://paste.ubuntu.com/916337/
[17:57] <zul> utlemming: http://paste.ubuntu.com/916324/ is the console-log, this is running on openstack as well
[17:59] <utlemming> zul: can you get me the output of "ec2metadata" (scrub appropriately)
[18:01] <uvirtbot> New bug: #912267 in python-keystoneclient (universe) "[MIR] python-keystoneclient" [Undecided,Fix released] https://launchpad.net/bugs/912267
[18:03] <zul> utlemming: http://paste.ubuntu.com/916349/
[18:07] <utlemming> zul: at face value, this is a bug in cloud-init
[18:07] <zul> utlemming: yeah
[18:32] <zul> utlemming: http://paste.ubuntu.com/916324/
[18:32] <zul> doh..
[18:32] <zul> utlemming: https://bugs.launchpad.net/ubuntu/+source/cloud-init/+bug/974509
[18:32] <uvirtbot> Launchpad bug 974509 in cloud-init "Issues resolving ubuntu archives." [Undecided,New]
[18:37] <hallyn> jdstrand: in http://people.canonical.com/~jamie/libvirt/qatest.tar.bz2 for qa-regression-testing, the qatest.xml really shouldn't have <emulator>/usr/bin/qemu</emulator>
[18:37] <smoser> zul, use opendns or google dns
[18:37] <smoser> 8.8.8.8
[18:38] <zul> smoser: yeah
[18:38] <smoser> i did know of this potential issue whe i did this.
[18:38] <smoser> dns redirectors suck
[18:40] <smoser> oh. shoot. you're not even hitting the bug i programmed for.
[18:40] <smoser> you can pass cloud-config to specify apt_mirror and it wont do that.
[18:41] <smoser> or you can fix your dns.
[18:41] <uvirtbot> New bug: #974509 in cloud-init (main) "Issues resolving ubuntu archives." [Undecided,New] https://launchpad.net/bugs/974509
[18:41] <utlemming> smoser: he's canadian
[18:41] <utlemming> and that means he has rogers isp, which is know for crappy DNS games
[18:42] <smoser> as are others.
[18:42] <smoser> but you can set what you use.
[18:42] <smoser> 8.8.8.8 will not give you such bogus stuff
[18:42] <smoser> neither will opendns
[18:44] <utlemming> opendns does now...I stopped using them becuase of that
[18:45] <utlemming> http://paste.ubuntu.com/916401/
[18:45] <jdstrand> hallyn: re the emulator line> how come?
[18:46] <hallyn> jdstrand: bc /usr/bin/qemu doesn't exist any more
[18:47] <hallyn> though maybe we have to explicitly use qemu-system-i386
[18:47] <jdstrand> test_emulators() in test-libvirt.py will need to be adjusted as well
[18:48] <hallyn> dat
[18:48] <hallyn> drat
[18:48] <hallyn> i don't suppose that could be related to test_guest_aa_attach_detach_physical hanging
[18:49] <jdstrand> I would expect massive failures if the emulator was wrong
[18:50] <jdstrand> hallyn: this is going to have to be changed in a way that is compatible going to lucid. one way to do that would be to conditionally update the xml on 12.04 and higher
[18:50] <peter_> hi all.. i got a bunch of servers running ubuntu all giving disk speeds of 30mb/s then i noticed that installs with my newer template gives 130mb/s.. just can seem to find the reason behind this. all are running the same ubuntu server version+kernel. must be some module or config? they are xenserver virtualized. any pointers/help appreaciated
[18:51] <jdstrand> hallyn: that could be done in LibvirtTestCommon::_setUp()
[18:51] <hallyn> jdstrand: the <emulator> line really shouldn't ever have been all that important...  qemu-kvm picks /usr/bin/kvm if it exists
[18:51] <hallyn> maybe _setUp() should just create the qemu symlink :)
[18:51] <jdstrand> hallyn: well, maybe... but this dates back to karmic
[18:52] <jdstrand> hallyn: yeah, that could happen too
[18:55] <hallyn> jdstrand: any advice on test_guest_aa_attach_detach_physical() hanging?
[19:08] <jdstrand> hallyn: it hangs forever? there is a 30 second sleep in there
[19:09] <hallyn> jdstrand: no, virsh -c qemu:///system attach-interface qatest-i386 network default --mac 52:00:00:00:00:00  actually hangs forever
[19:09] <hallyn> i was thinking it was the script, but it's just the virsh action hanging
[19:09] <hallyn> checking git log...
[19:09] <jdstrand> hallyn: that was going to be my next suggestion (checking virsh)
[19:12] <Daneil54> hi
[19:13] <hallyn> recon 'virNetSocketReadWire:996 : End of file while reading data: Input/output error' is involved
[19:13] <hallyn> actually maybe this is where the command.c patch i was going to put in but hadn't yet would help.
[19:13] <jdstrand> exit
[19:13] <jdstrand> heh
[19:13] <Daneil54> i have a probleme with subversion
[19:13] <Daneil54> can someone help please
[19:13] <Daneil54> ?
[19:21] <hallyn> Daniel54: I'd recommend looking for a #subversion
[19:39] <tash> shouldn't setting root: test@test.com in /etc/aliases mean that an crons that email root should be sent to test@test.com?
[19:58] <hallyn> feh, how many times does it re-run the same (failing) attach-inteface test?
[20:19] <chmac> Rapid2214: Apologies, got disconnected earlier, thanks again for your help, it seems to a be a mysql specific issue so I'll start investigating in that direction :-)
[20:19] <chmac> Think I got dropped before that got through a few minutes ago :-)
=== bladernr_ is now known as bladernr_afk
=== bladernr_afk is now known as bladernr_
=== bicranial_ is now known as bicranial
[21:21] <adam_g> smoser: ping
[21:22] <adam_g> dah, nvm
[22:10] <hallyn> jdstrand: actually, when I connect to qatest-i386 with vnc, it seems the problem is that it doesn't boot.  just hangs at 'uncompressing linux'
[22:11] <hallyn> i think the attach-interface is just hanging while qemu waits for the os to ack it
[22:11] <hallyn> i guess 'booting hte kernel' is where it hangs
[22:15] <hallyn> hm, i see.  seems perhaps a regression with unaccelerated qemu
[22:22] <tash> this is really strange, has anyone ever had this happen? Boot up and you've got a black screen with a blinking cursor?
[22:23] <tash> After a few reboots, eventually able to login, then after a few moments, things locked up again .... maybe SATA controller going bad or hard drive issue?
[22:25] <adam_g> zul: gonna start populating this: https://wiki.ubuntu.com/Openstack/Essex/Notes
=== Arc_ is now known as a5m0
=== matsubara is now known as matsubara-afk
[23:24] <RoyK> tash: check dmesg
[23:24] <RoyK> tash: or setup syslog logging to another host
=== Lcawte is now known as Lcawte|Away