[01:18] <pmatulis> ash_m: open a bug if something is unclear - https://bugs.launchpad.net/serverguide/+filebug
[01:22] <jvwjgames> HI i need help
[01:22] <jvwjgames> I can't Update anymore
[01:23] <jvwjgames> error read this
[01:23] <jvwjgames> Encountered a section with no Package: header
[01:23] <jvwjgames> Problem with MergeList /var/lib/apt/lists/us.archive.ubuntu.com_ubuntu_dists_trusty_multiverse_i18n_Translation-en%5fUS
[01:23] <jvwjgames> The package lists or status file could not be parsed or opened.
[01:23] <jvwjgames> Can't call method "policy" on an undefined value at /usr/bin/apt-show-versions line 56.
[01:23] <jvwjgames> E: Problem executing scripts APT::Update::Post-Invoke-Success 'test -x /usr/bin/apt-show-versions || exit 0 ; apt-show-versions -i'
[01:23] <jvwjgames> E: Sub-process returned an error code
[01:23] <jvwjgames> root@jvwjgames:~# rm /var/lib/apt/lists/us.archive.ubuntu.com_ubuntu_dists_trusty_main_i18n_Translation-en%5fUS
[01:26] <jvwjgames> nevermind
[01:26] <jvwjgames> i fixed it
[01:26] <jvwjgames> if any has this problem use this command rm /var/lib/apt/lists/* -vf
[01:26] <jvwjgames> and it will be fixed
[01:27] <jvwjgames> you just need to do an update
[01:27] <jvwjgames> an apt-get update that is
[01:30] <ash_m> pmatulis: do you think it is unclear? -- is this an issue where I am just too lacking in experience to understand that part of the documentation?
[08:08] <svetlana> How to upgrade from 14.04 to 14.10? do-release-upgrade does not pick it up.
[08:09] <Sling> svetlana: what happens when you do that though?
[08:09] <svetlana> Sling: it says it does not see any next release.
[08:11] <Sling> svetlana: what does Prompt contain in /etc/update-manager/release-upgrades
[08:11] <Sling> if you are currently not on an LTS release (which 14.04 isn't) and your Prompt is 'lts' then it won't upgrade, afaik
[08:12] <svetlana> Sling: "lts"
[08:12] <Sling> so setting it to 'normal' would fix that
[08:12] <svetlana> Ok. Ta.
[08:12] <Sling> you can always put it back to 'lts' afterwards to avoid getting non-lts releases later
[08:41] <lordievader> Good morning.
[09:37] <YamakasY> good morning!@
[09:38] <YamakasY> whoops
[09:38] <YamakasY> is this channel more alive again ?
[09:42] <lordievader> YamakasY: More alive? Compared to when?
[09:43] <YamakasY> lordievader, some weeks ago
[09:43] <lordievader> YamakasY: Its allways about this active. It is nice and quiet, everything works ;)
[09:45] <YamakasY> lordievader, hehe, yes better than #ubuntu, I got banned some days ago as I´m not there that much and someone pinged me about Ubuntu finally running from on a rasp, those ops were freaking out it was untrue, we were forcing together and baf..
[09:45] <YamakasY> strange channel these days
[09:46] <YamakasY> IRC changed :(
[09:52] <lordievader> Not really.
[09:52] <lordievader> In my opinion anyways.
[09:59] <YamakasY> lordievader, the #ubuntu channel is, it was dead for a while now we have ops there that want to answer a question before it´s even asked... and all stuff is most n00bish
[10:06] <svetlana> YamakasY: it is not like irc changed. More like irc channel.
[10:06] <YamakasY> svetlana, yap true
[10:06] <YamakasY> svetlana, you name :P I must ask are you a male or female ?
[10:07] <YamakasY> *your
[10:07] <svetlana> I manage to get answers when I want to still. What leaves me worried is that the distro contributes non-free stuff quite a bit. But I like spreading knowledge anyway and the way the channel is does not /get in the way/ of me doing so, so I'm happy.
[10:07] <svetlana> I'm female.
[10:08] <YamakasY> that is good! female in IT, finally not some male that takes a female name on iRC
[10:08] <svetlana> Waiting on an upgrade to 14.10 now. It includes systemd by default, which is forced on me during the upgrade. Hopefully it boots; in that case I can figure it out.
[10:08] <svetlana> I do not do IT as a job, only as a hobby.
[10:08] <YamakasY> oh they are changing too now ?
[10:08] <YamakasY> okay
[10:08] <YamakasY> good hobby
[10:08] <svetlana> I saydo not do a job in IT if you can. Most of the jobs result in proprietary output.  Consider non-IT job and do IT as a hobby.
[10:08] <YamakasY> #ubuntu is really for hobby now it seems, the channel lacks
[10:09] <YamakasY> I asked once something, #debian is better in those cases
[10:09] <svetlana> I find it okay to ask in #ubuntu, it helps others to learn.
[10:09] <svetlana> If they don't know, I reformulate the question, and eventually figure it out.
[10:10] <svetlana> There was no Ubuntu question I had which would be impossible to address with the help of #ubuntu and the web, thus far.
[10:10] <YamakasY> yeah but the way is wrong, the ops and some tard think they are king... they are just a number on this world to be honest
[10:10] <YamakasY> you can google everything about Ubuntu these days
[10:10] <YamakasY> much better
[10:12] <svetlana> I think it is not as much problem with ops as persons. It is a problem of the channel being too big and too fussed about offtopic stuff.  If one invents a way to split it, it'll become okay. And yesh, more people helping in there would also help.
[10:13] <svetlana> I am about to fall asleep. Through this dog slow internet and 1GB of upgrades.
[10:13] <YamakasY> haha
[10:15] <lordievader> Utopic does not run systemd by default, it has it installed due to udev.
[10:15] <lordievader> Even 15.04 doesn't run systemd by default.
[10:17] <YamakasY> okay, so it won´t be default ?
[10:17] <YamakasY> ever ?
[10:17] <YamakasY> I mean why do they want to switch ?
[10:19] <Odd_Bloke> YamakasY: http://www.markshuttleworth.com/archives/1316 is sabdfl's blog post on the matter.
[10:20] <YamakasY> Odd_Bloke, thanks! will read that after I fixed my primary key issue for MySQL
[10:31] <svetlana> ah
[10:31] <svetlana> now i should probably go learn what udev is
[10:35] <lordievader> svetlana: Udev is the replacement for HAL. It's a hardware abstraction layer.
[10:37] <YamakasY> oh yeah the Hal days and Videocards :S
[11:11] <jamespage> rbasak, I find my memory around library soname bumps slightly hazy? do you have a moment to double check something for me?
[11:11] <rbasak> jamespage: sure
[11:11] <jamespage> rbasak, so the new corosync point release for vivid bumps libvotequorum from 6->7
[11:11] <jamespage> fine
[11:12] <jamespage> so I think that all I need todo is cut the new libvotequourum7 package, drop the 6 one and then update deps for transition support
[11:12] <jamespage> although other than corosync, nothing else depends on this...
[11:12] <rbasak> Ah, I hadn't addressed corosync yet. Thank you for looking at it.
[11:12]  * rbasak is still working on MySQL :-/
[11:13] <jamespage> rbasak, keep working in mysql :-)
[11:13] <jamespage> rbasak, oh I took some time on my travels to look at dpdk as well
[11:15] <rbasak> jamespage: agreed. I don't think you need to do anything else. As long as that's the only sover bump?
[11:15] <jamespage> rbasak, as far as I can see yes it is
[11:16]  * rbasak wonders why they're shared libraries in the first place
[11:16] <rbasak> I suppose there might be out-of-archive builds against it.
[11:17] <rbasak> Doesn't sound exactly like a point release though! :)
[11:17] <jamespage> rbasak, yeah I was a little surprised
[11:17] <jamespage> probably not a candidate for a MRE :-)
[11:18] <rbasak> I suppose it's better that they bumped the sover than breaking ABI and not bumping anything at all.
[11:18] <rbasak> Not that it would impact anything in this case!
[11:22] <jamespage> rbasak, http://paste.ubuntu.com/10140929/
[11:22] <jamespage> does that look sane?
[11:30] <rbasak> jamespage: you might not need to Conflict/Replace libvotequorum5 any more as I don't see it in the archive.
[11:31] <jamespage> rbasak, good point
[11:31] <rbasak> jamespage: looks fine. There aren't any symbols files that need updating, do they?
[11:31] <jamespage> rbasak, so just 6 required for upgrades - makes sense
[11:31] <jamespage> rbasak, nope - checked that already
[11:31] <rbasak> jamespage: I would prefer to test the binaries though - I guess you know how to do that better than me.
[11:32] <rbasak> A dep8 test would be nice :-)
[11:32] <jamespage> rbasak, a bit tricky with a single instance
[11:32] <rbasak> Yeah. I wonder if we should start arranging amulet tests for this kind of thing.
[11:32] <jamespage> rbasak, but even a basic package install test is useful
[11:32] <rbasak> kickinz1 has a similar need with a postgresql clustering bug he's working on.
[11:32] <rbasak> In all that free time you have, of course :)
[11:33] <jamespage> rbasak, gnuoy has been working on that sort of stuff using mojo - we have the start of a full openstack ha cloud scenario
[11:33] <jamespage> kills units, makes sure stuff is still functional etc...
[11:33] <rbasak> That sounds good.
[11:34] <rbasak> Now that we have generally good dep8 coverage or at least a plan where that's missing, I think this might be the next testing hole for general server packaging.
[12:44] <coreycb> zul, jamespage: juno stable is all set for review.  cinder, neutron, and nova are dependent on this though - https://bugs.launchpad.net/ubuntu/+source/nova/+bug/1419117
[13:14] <Piffer_> In order to have crontab (root) execute a script, that stops a app, deletes a file, and starts the app again. When I run the script manually as root, it works fine. When cron runs it, the app is stopped, file deleted, but the app is not started. Are there any special tricks to have cron start an app?
[13:15] <maswan> No, but the environment might be a bit different. Maybe something is not in the path when run from cron or something needs to be set?
[13:18] <Piffer_> thanks, I think its a path issue.... that makes sense.
[13:25] <Piffer_> maswan: Thanks, it was a path issue. I had added the folder to the path, but not the sub folder containing the config file.
[13:28] <jamespage> coreycb, looking at you eventlet bump now
[13:32] <jamespage> coreycb, made one minor tweak
[13:32] <jamespage> convention is to use a point on the packaging version
[13:32] <jamespage> 1ubuntu4 -> 1ubuntu3.1
[13:32] <jamespage> tbh it did not really matter as vivid has a new upstream release
[13:32] <jamespage> coreycb, are you good on how to specify that as the min version for the things that need it?
[13:33] <coreycb> jamespage, did you mean 1ubuntu4 -> 1ubuntu4.1 ?
[13:33] <jamespage> coreycb, no
[13:34] <jamespage> the current utopic release version is 1ubuntu3 right?
[13:34] <coreycb> jamespage, ah, right..
[13:34] <jamespage> so just add a minor version bump
[13:35] <coreycb> jamespage, got it. so is there a trick in debian/control for packages that use eventlet?  currently cinder has python-eventlet (>= 0.13.0)
[13:36] <jamespage> coreycb, as you want to ensure that you get the right patch in packaging you need to extend taht
[13:36] <jamespage> (hint: and deal with the cloud archive backport at the same time)
[13:38] <coreycb> jamespage, something like... python-eventlet (>= 0.13.0-1ubuntu3.1~)
[13:41] <jamespage> coreycb, +1 spot on
[13:45] <coreycb> jamespage, Ok I'll make the updates to the openstack packages.  any idea when eventlet will be available for me to rebuild?
[13:46] <jamespage> coreycb, well I suspect the SRU team will process all-in-one go
[13:46] <jamespage> coreycb, so don't depend on it being in the proposed archive just yet - you'll need to work from a ppa for now I guess
[13:47] <coreycb> jamespage, ok
[13:52] <jamespage> rbasak, that corosync update works OK - although it did reveal a slightly nasty side effect of the upgrade
[13:53] <jamespage> pacemaker needs a restart with the newer corosync version
[13:54] <rbasak> jamespage: interesting. I suppose the right way would be arrange that inside postinsts somehow (not sure if that's allowed by policy directly though). But maybe not worth it?
[13:54] <jamespage> rbasak, I'm actually pondering whether this is another ceph like think
[13:54] <rbasak> I don't follow. ceph like thing?
[13:55] <jamespage> rbasak, for ceph, the restart of daemons must be co-ordinated across a cluster, otherwise you get outage
[13:55] <rbasak> I see
[13:55] <jamespage> rbasak, so the package maintainer scripts don't stop/start anything during upgrades; that's up to the operator
[13:55] <jamespage> rbasak, its still not ideal but its more controllable
[13:55] <rbasak> Makes sense.
[13:55] <jamespage> rbasak, the same might apply to a clustered system set
[13:56] <jamespage> you want to upgrade the binaries, then restart the daemons in a controlled way
[13:56] <rbasak> I suppose corosync/pacemaker is fundamentally a multi-server thing that packaging can't be expected to orchestrate.
[13:56] <rbasak> So more controllable, with the packaging doing less, makes sense.
[13:59] <jamespage> rbasak, I'll not poke at that for now
[14:00] <jamespage> rbasak, tbh I think some of its due to the version bump which won't happen post release
[14:01] <rbasak> ack
[14:10] <hariom> Hi, I am using virtualbox for running Ubuntu 12.04 server edition. The command line screen is very small despite install guest addition. How to increase size of the command line screen?
[14:23] <CiPi> Hello, i have a question, i have upgradedfrom 12.04 to 14.04, everything worked fine, but the only problem is that is booting on the old kernel, and how can i boot from kernel 3.13 ?
[14:23] <CiPi> How can i boot from another kernel from terminal via ssh
[14:25] <coreycb> jamespage, you mentioned something earlier about  "it did not really matter as vivid has a new upstream release".   what did you mean?
[14:26] <jamespage> coreycb, oh - so the convention is to drop a minor point in - 3.1 for example
[14:27] <coreycb> jamespage, right, I got that part -- but this is for juno-utopic so not sure what the relation to vivid was
[14:27] <jamespage> coreycb, this would ensure that you would not conflict version wise with a pure packaging update in the later ubuntu release
[14:27] <jamespage> utopic->vivid
[14:27] <jamespage> upgrade path
[14:27] <jamespage> coreycb, but vivid has 0.15 so its fine
[14:27] <coreycb> jamespage, ok got it, thanks
[14:37] <hariom> http://www.mikeplate.com/2011/11/08/change-virtual-console-resolution-on-ubuntu/
[14:44] <CiPi> How can i boot from another kernel from terminal via ssh ?
[14:44] <CiPi> I have kernel 2.6, and installed kernel 3.13
[14:44] <CiPi> And after reboot is booting still from the old kernel 2.6
[14:47] <CiPi> Welcome to Ubuntu 14.04.1 LTS (GNU/Linux 2.6.32-042stab092.2 x86_64)
[14:47] <CiPi> Damn :|
[14:49] <rberg_> is that a VM getting its kernel from the host system?
[14:50] <teward> CiPi: is the system a VPS or something?
[14:50] <CiPi> Yes, is VPS.
[14:50] <teward> CiPi: do you know if it's OpenVZ or not?
[14:51] <CiPi> Yes, openvz
[14:51] <teward> CiPi: that's why it won't update - it's likely inheriting from the host
[14:52] <CiPi> Damn. :|
[18:36] <dannf> hallyn: hey, not sure what your plans are for qemu in vivid, but if you're sticking w/ 2.1 would you consider this? https://github.com/dannf/pkg-edk2/commits/armvirt
[18:48] <hallyn> dannf: i'm merging 2.2 this week
[18:49] <dannf> hallyn: cool, then you should already have all those changes
[18:49] <hallyn> dannf: if those changes unblock things pls feel free to push in the meantime though
[18:51] <dannf> hallyn: nah, if you're confident about a 2.2 this week, i can wait. thx :)
[18:54] <hallyn> dannf: i've got a ppa build of 2.2 so i don't *expect* any blockers
[18:57] <dannf> hallyn: ah, where can i find that?
[19:00] <hallyn> dannf: https://launchpad.net/~ubuntu-virt/+archive/ubuntu/virt-daily-upstream
[19:00] <hallyn> (getting a bit crufty by this point i guess)
[19:00] <dannf> thx
[19:18] <coreycb> zul, I'm working on kilo MIRs and enabling tests for pylibmc.  the version is pylibmc (1.2.3-1build1).  should I change that to utopic?
[19:18] <coreycb> sorry, to ubuntu
[19:19] <zul> coreycb: yes
[19:19] <coreycb> zul, ok
[21:12] <soahccc> Can someone confirm my suspicion that this is the network driver? https://gist.github.com/2called-chaos/c800bdb2421d95c30b12
[21:23] <GeekDude> soahccc: looks like a gist to me
[21:24] <GeekDude> It says "WARNING at /build/buildd/linux-lts-saucy-3.11.0/net/ipv4/tcp_input.c"
[21:24] <GeekDude> I'm no expert, but it looks like the tcp stack to me
[21:25] <soahccc> GeekDude: That's how I came to my suspicion ;)
[21:31] <mfisch> zul: is there a reason you dont put a Label: on the release file for UCA?
[21:31] <mfisch> reference: http://ubuntu-cloud.archive.canonical.com/ubuntu/dists/trusty-proposed/kilo/main/binary-arm64/Release
[21:31] <mfisch> it would make my pinning much easier
[21:32] <med_> we need bits of kilo and juno
[21:32] <mfisch> what med said
[21:37] <med_> http://us.archive.ubuntu.com/ubuntu/dists/trusty-updates/main/binary-amd64/Release
[21:37] <med_> mfisch, ^
[21:37] <med_> zul ^
[21:37] <med_> vs
[21:37] <med_> http://ubuntu-cloud.archive.canonical.com/ubuntu/dists/trusty-updates/kilo/main/binary-amd64/Release
[21:39] <mfisch> med_: here's the man page that talks about using labels: http://ccrma.stanford.edu/planetccrma/man/man5/apt_preferences.5.html
[21:41] <med_> maybe jamespage knows how to get OVS from UCA/kilo while running UCA/juno
[21:41] <med_> without just wgetting it (which does work fine)
[22:03] <zul> mfisch: ermm...not sure why we dont