[05:11] <andy7534211> so i made a PPA, but it was the first one i've made and it got all goofed up
[05:11] <andy7534211> so then i deleted it so i could re-create it from scratch
[05:12] <andy7534211> except launchpad kind of acted like i deleted it, but really didn't because it still shows up and i can't create a new one with the same name?
[05:13] <andy7534211> is there any way to *actually* delete it, or does it do that automatically after a week or so?
[05:15] <maxb_> andy7534211: No... it never gets deleted. Sorry.
[05:16] <wgrant> PPA deletion as it stands now exists mainly to let people rename their accounts.
[05:16] <wgrant> It could be taken further, but there wasn't an urgent need.
[05:16] <wgrant> And it's a lot harder to do a full deletion.
[05:18] <andy7534211> hm, and no way to rename or move the deleted ppa?
[05:19] <wgrant> Sadly not. It would be easy to automatically rename them to something else upon deletion, if you file a bug requesting that.
[05:19] <andy7534211> alright
[05:20] <wgrant> It's all pretty complicated, as we have to deal with stuff on disk under ppa.launchpad.net, as well as copies to and from other PPAs.
[05:20] <wgrant> So unrestricted deleting/renaming is problematic.
[05:20] <wgrant> But once the deletion is complete, we can automatically rename to something that's unlikely to conflict.
[05:22] <andy7534211> file against "Launchpad itself"?
[05:22] <wgrant> That's the one.
[05:22] <wgrant> https://launchpad.net/launchpad/+filebug
[05:28] <andy7534211> wgrant: would it also be easy to just recreate/reactivate the ppa?
[05:29] <wgrant> andy7534211: Ummm, sort of. It will work if you add /+edit to the URL and turn the Publish and Enabled flags back on.
[05:29] <wgrant> It's not supported, but it will work.
[05:30] <andy7534211> yikes, that was easy.. i think
[07:58] <mrevell> Hi
[09:13] <dpm> morning henninge, could you give me a hand with language packs? It seems that the Oneiric language pack exports from last Thursday and yesterday didn't happen -> https://translations.launchpad.net/ubuntu/oneiric/+language-packs
[09:17] <dpm> or jtv? ^
[09:18] <jtv> dpm: I'm more or less committed for the rest of the day I'm afraid.  :/
[09:18] <dpm> jtv, no worries, thanks
[09:19] <dpm> if anyone else could give a hand, that'd be great. We're nearing the second beta and it would be good to have working language packs
[09:25] <henninge> dpm: did you request a full language pack?
[09:25] <wgrant> henninge, dpm: Do you know when it was running?
[09:26] <wgrant> If it takes several or 24 hours, that's going to be difficult to do except over weekends.
[09:26] <henninge> wgrant: it starts at 22 utc
[09:26] <henninge> because of the roll-outs?
[09:26] <dpm> wgrant, henninge, I did not request a full export (the checkbox is not ticked on that page ^)
[09:26] <wgrant> Right, the DB disappears for 2 minutes every couple of days.
[09:28] <dpm> the two exports that are missing are delta exports, which do not take as near as long as full exports (which iirc take ~16 hours)
[09:28] <henninge> dpm: it does not really matter, the db reloads will kill it anyway.
[09:28] <wgrant> Ah, I see two failed langpack runs here...
[09:28] <wgrant> Let me see..
[09:29] <henninge> yes, just found them, too.
[09:29] <wgrant> DisconnectionError: server closed the connection unexpectedly
[09:29] <wgrant> As expected.
[09:29] <wgrant> Interesting.
[09:30] <wgrant> Looks like they'd been going for 16.5 hours?
[09:31] <henninge> dpm, wgrant: the last full export was quite a while ago
[09:31] <henninge> so, the delta will be quite large.
[09:32] <wgrant> So, if you'd requested one yesterday it would probably have been OK, since we (unpredictably) didn't have downtime tonight.
[09:32] <wgrant> If they take less than 48 hours, the weekend should be safe...
[09:33] <henninge> the last deta only took 8 hours
[09:33] <henninge> wgrant: they are on a schedule
[09:34] <henninge> dpm: can you please file a bug to make the language pack export restartable? It needs to be able to pick up where it left off.
[09:34] <henninge> Not sure that is easily possible, though.
[09:36] <dpm> wgrant, exports happen automatically on a cron job - https://dev.launchpad.net/Translations/LanguagePackSchedule. Full exports need to be requested through that page ^, but are also started according to the schedule
[09:36] <dpm> henninge, sure, give me a minute. However, what changed recently? Scheduled delta exports used to work until last week at least
[09:37] <henninge> dpm: last week we started to restart the database a few times a week.
[09:37] <henninge> wgrant: but we have a fairly fixed time for when that happens, right?
[09:38] <henninge> jtv: do you remember the specific reason to star language pack export at 22 UTC?
[09:38] <dpm> hm, so what can we do to ensure the langpacks are exported in the meantime, assuming that the bug to restart langpacks might take a while to get fixed?
[09:38] <henninge> I think the quickest solution would be to have them start right after the database restart window.
[09:38] <jtv> henninge: it used to be later, but we moved it forward a bit to ensure that it completed before langpack-o-matic runs.
[09:39] <henninge> ok, but we could move langpack-o-matic, too
[09:39] <jtv> We can if needed, yes, but it involves coordinating across teams.
[09:39] <henninge> jtv, dpm: when and where does langpack-o-matic run? Who controls that?
[09:39] <wgrant> henninge: 0830UTC
[09:40] <dpm> henninge, it's on the schedule ^
[09:40] <wgrant> henninge: That's fastdowntime.
[09:40] <henninge> right
[09:40] <wgrant> The window is 0830-0835
[09:40] <wgrant> We do occasionally run a minute over.
[09:40] <dpm> henninge, pitti or myself can adjust the langpack-o-matic runs
[09:40] <henninge> so if we start at 9UTC and the runs take under 23 hours, we are safe.
[09:40] <henninge> wgrant: that's ok ;-)
[09:41] <wgrant> "The timing is chosen so we hit the times when we usually expect the lowest load. Most of the translation work happens in Europe, South America and Asia, and 22:00 UTC should mean less load on LP due to translation activity. With DB slaves, though, this is probably not something we should worry about."
[09:41] <wgrant> Given that we are well under DB capacity, I agree with the last sentence.
[09:42] <dpm> wgrant, henninge - ok, so you suggest we should move the start times for the language pack export jobs in LP and adjust the langpack-o-matic builds accordingly?
[09:43] <wgrant> That would be my recommendation, but I'd also strongly suggest prioritisation of optimisations :)
[09:43] <henninge> yes, that is the only short-term solution I can see.
[09:43] <henninge> wgrant: come again? ;)
[09:43] <wgrant> henninge: You have two days left, you can do it!
[09:43] <henninge> ha!
[09:43] <nigelb> heh
[09:44] <nigelb> 48 hours should be just about enough :P
[09:49] <henninge> dpm: I will have to relocate now but will go about moving the jobs when I am back online.
[09:50] <dpm> thanks henninge, here's bug 849829
[09:52] <dpm> wgrant, could you confirm if the langpack export started yesterday at 22:00UTC is still running?
[09:52] <wgrant> dpm: If one started, it should be running.
[09:52] <nigelb> Wasn't there downtime today?
[09:52] <wgrant> We didn't interrupt them 1.5 hours ago.
[09:52] <wgrant> No.
[09:52] <wgrant> buildbot was broken, so we couldn't deploy.
[09:52] <wgrant> I don't have a failure email from yesterday's, either.
[09:52] <wgrant> So it is probably still running.
[09:52] <wgrant> a LOSA could confirm.
[09:53] <dpm> how do I get a losa to confirm?
[09:54] <wgrant> LOSA ping: could you please check if there's a language-pack-exporter.py from yesterday still running on loganberry?
[09:54] <jtv> wgrant: you're thinking #-ops
[09:54] <wgrant> Oh?
[09:54] <wgrant> LOSAs live here too, and taking the discussion to -ops hardly seems productive.
[09:55] <mthaddon> wgrant: no lang pack export running on loganberry or ackee
[09:55] <wgrant> Hm.
[09:55] <wgrant> Maybe it finished.
[09:55] <wgrant> I don't have an error email.
[09:55] <wgrant> Thanks mthaddon.
[09:57] <dpm> thanks mthaddon
[09:57] <mthaddon> np
[09:57] <dpm> wgrant, not sure it finished, it's not listed in there: https://translations.launchpad.net/ubuntu/oneiric/+language-packs
[10:00] <wgrant> dpm: http://paste.ubuntu.com/689029/
[10:00] <wgrant> Yesterday's run was for natty.
[10:02] <dpm> wgrant, ah, yes I confused the build day with the export day. I was thinking of Oneiric, so it seems the Oneiric export on Monday was interrupted and didn't happen
[10:05] <wgrant> dpm: Right, there have been interruptions at 0830UTC every weekday since last Thursday, except today.
[10:06] <wgrant> Amusingly enough, today was the one day it would have been OK.
[10:06] <dpm> oh well...
[12:44] <Sweetsha1k> is there an easy way to query launchpad for an upstream bug (i.e. 'also affects project')
[12:45] <Sweetsha1k> ?
[14:18] <abentley> matsubara: branch scanner oopes aren't showing up in the oops viewer, e.g. OOPS-2076SMS9, OOPS-2075SMS15, OOPS-2075SMS17
[14:19] <matsubara> abentley, yep, danilos told me about it yesterday. I'm still debugging the issue and filed https://bugs.launchpad.net/oops-tools/+bug/849446 for it
[14:19] <abentley> matsubara: great.
[14:20] <abentley> matsubara: you know that they moved the branch scanner to ackee recently, right?
[14:21] <matsubara> abentley, nope. that shouldn't be a problem though. it's likely there's a bug in the dir_finder script which is making it fail to find the new directories
[15:15] <danilos> matsubara, fwiw, could the problem be the fact there are "duplicate" directories for the same OOPS type (i.e. loganberry-bzr-syncd/branch-scanner and ackee-bzrsyncd/branch-scanner)?
[15:20] <matsubara> danilos, not sure. the dir_finder code isn't supposed to take that into account
[17:11] <deryck> over to you, abentley.....
[17:12] <abentley> deryck: ack
[21:59] <head_victim> Is there anyway to reset the quota for direct contact of other launchpad users?
[21:59] <head_victim> I'm contacting several users on behalf of a team and need to contact 4 more :/