[02:15] <spm> wgrant: sorry - did you get that log request thingy you were after sorted?
[02:16] <wgrant> spm: buildd-manager is really unwell.
[02:16] <wgrant> spm: Can you check if there's anything obvious in the logs?
[02:16] <spm> and in other news, water is wet.... ;-)
[02:16] <spm> sure
[02:19] <spm> oh yuk. obvious doesn't do it justice.
[02:19] <wgrant> Oh?
[02:20] <spm> wgrant: http://paste.ubuntu.com/470052/ oh.
[02:21] <spm> lots of 'em
[02:21] <wgrant> Um.
[02:21] <wgrant> Ummmm.
[02:21]  * wgrant screams.
[02:22] <wgrant> It doesn't give any indication of which build it is?
[02:22] <spm> hrm. wonder if that means I should be taking a sickie....
[02:22] <spm> not so far as I can tell
[02:22] <spm> no. lies. I chopped one line too early. openmotif.
[02:22] <wgrant> Do you have a build number?
[02:22] <wgrant> Or must we query around for it?
[02:22] <spm> startBuild(...., openmotif, 2.3.2-6, Release)
[02:22] <wgrant> OK, let's see..
[02:24] <spm> fgrep -B 1 "Scanning failed with: u'non-free'" /srv/launchpad.net/production-logs/buildd-manager.log <== same builder, same error, same package, every time.
[02:24] <wgrant> Can you tell which archive it's in? You should be able to see it in the sources.list entries.
[02:25] <spm> hrm. I think I need a little more context here?
[02:25] <spm> do you mean that file in /etc/apt?
[02:26] <wgrant> Sorry, the startBuild line should contain an 'archives' argument which contains a list of sources.list entries.
[02:26] <wgrant> That should tell us which archive is insane.
[02:26] <spm> Oh I see. right. looking.
[02:29] <wgrant> If not, we should be able to find it in the DB pretty easily.
[02:30] <spm> alas no. where the lines you're referring to appears in the log, we only get the traceback. :-/
[02:30] <spm> sql me
[02:31] <wgrant> OK. SELECT sourcepackagerelease.id, dateuploaded FROM sourcepackagerelease JOIN sourcepackagename ON sourcepackagename.id = sourcepackagerelease.sourcepackagename WHERE sourcepackagename.name = 'openmotif' AND version='2.3.2-6';
[02:32] <wgrant> Actually, get upload_archive too.
[02:32] <spm>  upload_archive |   id   |       dateuploaded
[02:32] <spm> ----------------+--------+---------------------------
[02:33] <spm>           14369 | 592626 | 2010-01-27 19:00:27.29329
[02:33] <spm> wgrant: ^^
[02:33] <mwhudson> uploaded in january?
[02:33] <spm> was a good month for uploads.
[02:33] <wgrant> Maybe it was copied...
[02:33] <mwhudson> six months + a few hours ago
[02:34] <mwhudson> i hope that this is a coincidence
[02:34] <spm> that wouldn't be a like a recipie build thing that thumper and I just enabled a few hours ago.....
[02:35] <spm> nah. can't be. that's 19:00; we enabled post 23:00.
[02:37] <wgrant> spm: SELECT binarypackagebuild.id, packagebuild.archive, binarypackagebuild.distro_arch_series, packagebuild.pocket, date_created  FROM binarypackagebuild JOIN packagebuild ON packagebuild.id = binarypackagebuild.id JOIN buildfarmjob ON buildfarmjob.id = packagebuild.build_farm_job WHERE binarypackagebuild.source_package_release = 592626;
[02:37] <wgrant> Er, that's not it.
[02:38] <wgrant> Ah, no, that is right.
[02:39] <spm> chanel spam enroute...
[02:39] <spm>    id    | archive | distro_arch_series | pocket |        date_created
[02:39] <spm> ---------+---------+--------------------+--------+----------------------------
[02:39] <spm>  1892335 |   19017 |                 81 |      0 | 2010-07-27 20:00:52.334521
[02:39] <spm>  1471081 |   14369 |                 81 |      0 | 2010-01-27 19:00:31.75819
[02:39] <spm>  1471082 |   14369 |                 82 |      0 | 2010-01-27 19:00:31.836872
[02:39] <spm>  1471083 |   14369 |                 84 |      0 | 2010-01-27 19:00:31.924909
[02:39] <spm> (4 rows)
[02:39] <spm> wgrant: ^^
[02:39] <wgrant> Oh look, a new build.
[02:39] <spm> dollars to pesos, if I check the back logs this errors starts @ approx 20:00 yesterday...
[02:40] <wgrant> We should probably just disable the archive.
[02:41] <wgrant> It has a lot of .... bad things.
[02:41] <wgrant> I have no idea how they got like that.
[02:41] <wgrant> But it has publications in non-free, even though PPA copies always override to main.
[02:41] <spm> perfectly matches. builddmanager starts logging this error: 2010-07-27 21:06:43+0100
[02:41] <wgrant> Hmm.
[02:42] <wgrant> spm: https://edge.launchpad.net/~jedimaster/+archive/ppa/+packages, expand the openmotify row. Is the archive it was copied from private?
[02:43] <wgrant> Oh, the team is actually owned by Brian.
[02:43] <wgrant> Oddd.
[02:43] <spm> and the answer to your first q; yes.
[02:44] <wgrant> OK, so I can't examine further. But something is seriously broken with that archive, and disabling it archive will get things running again until Soyuz appears.
[02:44] <wgrant> Ooh I wonder.
[02:44] <spm> how does one disable? rescore to -1? or am I missing something obvious. again.
[02:45] <wgrant> spm: Go to the Archive's +edit.
[02:45] <wgrant> It should be obvious from there.
[02:45] <wgrant> It will suspend all the builds, and is completely reversible by simply toggling the box again.
[02:46] <spm> oh! disable the entire archive. right. lalala.
[02:46] <wgrant> The archive is full of broken publications.
[02:46] <wgrant> And I think I know why.
[02:46] <spm> "This PPA has been disabled."
[02:46] <wgrant> Thanks.
[02:46] <spm> I'll send brian an email explaining why
[02:47] <wgrant> Yeah, I've found the bug.
[02:47] <wgrant> Damn.
[02:47] <wgrant> Delayed copies don't obey the PPA component override.
[02:48] <wgrant> Luckily those publications won't have been able to publish, so cleanup is fairly easy once Soyuz appears.
[02:51] <wgrant> And everything looks happy again.
[02:51] <spm> i like happy things
[02:52] <spm> cc'd jules on the email to brian; so he'll have some context to work with.
[03:03] <spm> wgrant: btw, once again, what alerted you to there being a problem? was is stuff just not moving in the builds summary page?
[03:03] <wgrant> spm: /builders was just about empty
[03:03] <spm> ahh, and it shouldn't be. right. ta.
[03:03] <wgrant> Right, the queues were large, but the builders were idle for... a long time.
[03:04] <spm> wgrant: I wonder how hard it would bee to write a nagios plugin that interfaced with yourself. any thoughts?
[03:04] <wgrant> Haha.
[03:07] <lifeless> this is why I want a fresh dashboard
[03:08] <lifeless> people get attuned
[03:09] <spm> for sure; but that also requires someone looking at same; what worries me is that things (not snigling soyuz out here, but the exmaple is current) break and we don't have checks to pick up on them breaking. And tbh, not really sure what a decent check would be that doesn't generate a bazzilion false alarms. Be Ware of the boy crying wolf.
[03:10] <lifeless> spm: exactly
[03:11] <spm> the goal of an alert should be - if this triggers a *person* MUST do something. if a person can't do something, the alert is pointless and merely adds noise.
[03:11] <wgrant> Bug #610687, anyway.
[03:11] <spm> ta
[03:11] <lifeless> spm: right; dashboard - used ones - can be better at that
[03:11] <lifeless> spm: what we can automated as alerts, obviously, we should
[03:12] <spm> or even email summaries - eg the oops reports. it's not a zomg critical; but this is important that someone needs to make a determination soon.
[03:12] <lifeless> they need a trim
[03:12] <spm> no idea; I don't read them. :-/
[03:15] <spm> actually - 'a person must do something now', slight update. ie drop everything else, and fix. So any and all RT's for whomever; they immediately take the backseat; kinda thinking.
[03:16] <lifeless> stop the line
[03:16] <spm> well; even more so.
[03:16] <lifeless> uhm
[03:16] <spm> it's not stop; it's switch into emergency response mode.
[03:16] <lifeless> thats what stop the line /means/ in lean terms
[03:16] <lifeless> something is wrong, fix it, gather data about cause, feed that into make-things-better-later
[03:17] <spm> I can see the simularity. but generally, no. emergency response covers it better. something critical is broken. unbreak it. yesterday.
[03:18] <lifeless> stop the line originates on the factory floor, where failing to stop the line == million dollars of wrecked parts, or whatever.
[03:18] <lifeless> it really is intended to be critical
[03:18] <lifeless> if its been diluted, thats a problem
[03:19] <spm> heh. DEC support had a classic example of a Sev1 call. "Hi, we have semi-trailers backing up down St Kilda Road. We need this fixed. Now."
[03:36] <cody-somerville> Is Soyuz back to normal now?
[03:36] <spm> cody-somerville: .... rather loaded question isn't it?
[03:37] <cody-somerville> well, are packages getting built?
[03:37] <spm> heh, looks like it.
[03:38] <wgrant> It'll take a while to clear.
[03:39] <wgrant> But yes, things seem to be operating OK now.
[03:39] <wgrant> Besides the barrage of private builds destroying the official builders.
[03:40] <cody-somerville> whats with all the idle builders in the PPA pool?
[03:42] <wgrant> It may be in the middle of a dispatch run.
[03:43] <wgrant> It looks like it is.
[03:43] <wgrant> We'll see in a couple of minutes if the dispatches have stuck.
[03:43] <wgrant> But it's looking OK.
[03:45] <wgrant> Yep, it's all good.
[03:45] <wgrant> Cycles are up to more than three minutes, but it's at least working.
[12:38] <yofel> is there a place someone can comment on the new malone layout on edge? IMHO the flames look kind of lost ...
[12:38] <yofel> https://bugs.edge.launchpad.net/ubuntu/+source/qtm/+bug/573785
[13:12] <X3> hi where can I find the link to https://code.edge.launchpad.net/~team-iquik/lmsensors/trunk in https://launchpad.net/~team-iquik
[13:12] <X3> team is set as owner yet nothing shows up on the team page relating to that
[13:19] <wgrant> X3: Click on the 'Code' tab at the top.
[13:20] <X3> ah cheers
[13:22] <X3> wgrant: that did it, though I haf expected that to be part of the description page
[13:23] <X3> though makes sense whwre it is
[13:36] <wgrant> X3: Teams can have loooots of bugs and branches and other things associated with them. Showing them all on the one page isn't always feasible.
[13:37] <X3> mk
[14:40] <bigon> hi
[14:40] <bigon> I've added the ubuntu-mozilla-daily ppa but it seems that some info of the Release file are not take into account by apt (like origin...)
[14:40] <bigon>  500 http://ppa.launchpad.net/ubuntu-mozilla-daily/ppa/ubuntu/ lucid/main Packages origin ppa.launchpad.net
[14:42] <bigon> is there something wrong in the release file generated? or a bug in apt
[14:44] <maxb> bigon: What is the problem exactly?
[14:45] <bigon> other ppa have a line like that when typing apt-cache policy
[14:45] <bigon>  500 http://ppa.launchpad.net/bigon/ppa/ubuntu/ lucid/main Packages release v=10.04,o=LP-PPA-bigon,a=lucid,n=lucid,l=PPA for Laurent Bigonville,c=main origin ppa.launchpad.net
[14:46] <bigon> but with the ubuntu-mozilla-daily it miss all the info from the release file
[14:46] <maxb> The information is there in the Release file, but I have noticed that apt sometimes seems to fail to read it
[14:48] <maxb> You could try deleting /var/cache/apt/*pkgcache.bin, and /var/lib/apt/lists/*_*, and running apt-get update again - maybe it'll then pick it up
[14:50] <bigon> deleting the pkgcache.bin did the trick
[14:51] <bigon> so bug in apt
[14:51] <bigon> ::
[14:51] <bigon> :/
[14:52] <bigon> thx
[15:46] <damoxc> is there any reason why a build recipe is going to take 2 days?
[15:58] <bigjools> damoxc: they're scored very low priority right now while we're beta testing, they'll get parity with normal builds soon and will be done quicker
[16:00] <pindonga> hi. I have a project where I don't get branches automatically marked as merged... may this have to do with the project not having a series registered?
[16:06] <damoxc> bigjools: ah okay, cool!
[18:19] <micahg> is there no longer a subscribe someone else to questions?
[18:25] <shadeslayer> hey, in daily recipes, how can i get the version to be fixed_version+{bzr rev no}~{release}~ppaz
[18:25] <shadeslayer> is {release} a variable?
[18:46] <shadeslayer> anyone around to help with https://code.edge.launchpad.net/~team-iquik/+recipe/lmsensors-trunk
[18:46] <shadeslayer> it seems our daily build archive is right
[18:46] <shadeslayer> but under the archive part it says  addons
[20:58] <gnomefreak> OOPS are getting annoying now
[20:58] <gnomefreak> Error ID: OOPS-1670ED4276 Error ID: OOPS-1670EC4084 just in the last 30-60 seconds
[21:12] <sshaw> I have a team where we have 2 PPAs.  I setup my .dput.cf with my personal ppa and then the 2 team ones.  When I tried to update to my 'stable' ppa all of the packages got my 'internal' ppa instead
[21:13] <sshaw> I'm not sure what I'm doing wrong.  Here is my .dput.cf - http://paste2.org/p/931010
[21:13] <sshaw> if anyone has any insights
[21:14] <bigjools> sshaw: what command line did you use?
[21:15] <sshaw> bigjools: originally dput stable file.changes
[21:15] <sshaw> last time, dput ppa:mono-a11y/ppa
[21:15] <sshaw> (hoping to push it to stable)
[21:16] <bigjools> I'd delete your dput.cf for starters
[21:16] <bigjools> it might be a bug in dput
[21:18] <sshaw> ok, deleted and tried to re upload the package
[21:19]  * sshaw finds it odd though that it didn't ask for a password
[21:20] <yofel> sshaw: you signed the package when creating it, iirc that's all it needs for authentification
[21:22] <sshaw> oh
[21:22] <sshaw> rejected
[21:22] <sshaw> it says that its already in ubuntu/karmic
[21:23] <sshaw> although deleted, there is a mono-uia package in queue to be built under internal, but nothing under stable
[21:50] <bdmurray> I just got oops-1670ea4372 when reporting a bug and the end of the traceback is in connection_raw_execute when trying to insert data into message
[21:53] <kirkland> is launchpad misbehaving right now?
[21:54] <kirkland> i can't get anything to load;  don't see a status page
[21:54] <bdmurray> kirkland: anything?
[21:55] <kirkland> bdmurray: restarted my browser, seems to be happier now
[21:55]  * kirkland eats his words
[21:55]  * kirkland shakes his finger at chromium
[21:55] <bdmurray> I was gonna say works for me!
[21:56] <bdmurray> well aside of that oops I got
[22:28] <ZandreBran> Hi I am getting an error when compiling the orca in the PPA. On my computer I can compile without probelmas. Please I need help. Anyone?
[22:29] <wgrant> ZandreBran: Have you tried to build it with pbuilder?
[22:30] <ZandreBran> warp10, I will check
[22:32] <ZandreBran> warp10, not. I'm using on my PC to create the .deb: debuild -us -ua
[22:33] <ZandreBran> warp10, for send PPA: debuild -S
[22:33] <wgrant> ZandreBran: You should always try building with pbuilder locally first, to ensure that your package builds in a clean environment.
[22:33] <wgrant> Once you've fixed it so it successfully builds in pbuilder, it will probably build on Launchpad.
[22:34] <ZandreBran> warp10, i'm try pbuilder now. thanks and sorry my english, living in Brazil.
[22:34] <micahg> ZandreBran: Tab Complete Failure :)
[22:35] <ZandreBran> wgrant, micahg, ops. sorry.
[23:26] <X3> hi this https://code.edge.launchpad.net/~team-iquik/+recipe/lmsensors-trunk
[23:27] <X3> says in 4 hours but theres no change to time after almost that much
[23:28] <wgrant> X3: Yeah, the build queue is a bit backed up at the moment, and recipe builds don't have the highest priority.
[23:28] <X3> i know their lower priority
[23:28] <X3> backed up ya it figures
[23:29] <X3> i had fuse waiting to build for near 5 hours
[23:29] <X3> on normal ppa
[23:30] <X3> so i guess the recipe is gonna be slower
[23:30] <X3> can anything be built via a recipe?
[23:31] <X3> e.g. xbmc from svn trunk?
[23:31] <wgrant> If you have a Launchpad code import of their trunk, sure.
[23:32] <X3> Im gonna test a setup, but wanted to know what sort of recipe I could use
[23:50] <X3> th svn I wanna import is http://xbmc.org/download/
[23:51] <X3> if I enter the trunk http it spews this out
[23:51] <X3> Launchpad has not been able to mirror this branch. The last attempt was 3 minutes ago. (Not a branch: "http://xbmc.svn.sourceforge.net/svnroot/xbmc/trunk/".) Launchpad will try again in 23 hours. If you have fixed the problem, please ask Launchpad to try again
[23:52] <X3> any idea why?
[23:52] <wgrant> X3: That sounds like you created a mirror (for bzr branches), when you needed to create an import.
[23:52] <wgrant> Let's see...
[23:52] <X3> ah
[23:53] <wgrant> There's already https://code.edge.launchpad.net/~vcs-imports/xbmc/upstream, but that's old and seems to be for a linuxport branch.
[23:53] <wgrant> Should it just be importing trunk now?
[23:53] <X3> na
[23:53] <X3> way too old
[23:53] <X3> it needs to be the live trunk
[23:54] <X3> under brach type thers 3 selections
[23:54] <X3> hosted mirror remote
[23:54] <X3> maybe remote is it
[23:54] <wgrant> It's none of those. It's a separate thing, because you're importing from Subversion.
[23:54] <wgrant> Let me find it.
[23:54] <X3> cheers
[23:55] <wgrant> X3: https://code.edge.launchpad.net/xbmc/+new-import
[23:56] <wgrant> You should just have to enter the Subversion URL there.
[23:57] <mtaylor> where did the thing I can use to merge user accounts run off to?
[23:57] <wgrant> mtaylor: https://launchpad.net/people
[23:57] <X3> wgrant your a star
[23:58] <wgrant> Now you just need someone to review it and get it started.
[23:58]  * thumper can do that
[23:59] <mtaylor> thanks wgrant