wgrantbigjools: "requested to review"00:00
wgrantIt's the reviewer00:00
wgrantNot a subscriber00:00
bigjoolsso that overrides sibscription settings saying not to email?00:00
bigjoolsok how do I stop this?00:00
lifelessit doesn't override, its direct notification00:00
lifelessremove maas-lander from maas-maintainers00:00
wgrantSet a team email address00:00
wgrantOr remove maas-lander from maas-maintainers, yeah00:00
lifelessput maas-lander and maas-maintainers both in ~canonical-launchpad-branches or a similar branch-only-access team00:01
bigjoolsthen it won't be able to land stuff00:01
mwhudsonlifeless: that's basically a way for evil js to get at http-only cookies isn't it?00:01
bigjoolsman this stuff is over complicated00:01
lifelessthats *why* we have that structure in LP, and why I encouraged you to use the same teams, or the same structure, for MAAS :)00:01
bigjoolslifeless: I understand - or rather I don't understand why it needs to be so complex :(00:01
lifelessmwhudson: http://en.wikipedia.org/wiki/CRIME_%28security_exploit%2900:01
mwhudsonlifeless: "It relies on the attacker being able to observe the size of the ciphertext sent by the browser while at the same time inducing the browser to make multiple carefully crafted web connections to the target site."00:02
mwhudsonlifeless: if you can do that, you can just carefully craft the results to do evil directly, surely?00:02
lifelessmwhudson: yes, but the demonstration is terrifying00:02
mwhudsonit's a super cool hack00:03
lifelessmwhudson: no, you can't00:03
wgrantmwhudson: No00:03
wgrantmwhudson: By crafting part of a body I can reveal the headers00:03
wgranteg. if I control a string embedded deep in JSON somewhere00:03
bigjools!!!!!! I can't set a contact address that is the same as another team's00:04
lifelessbigjools: I wouldn't go the contact address route00:04
lifelessbigjools: it gets super messy.00:04
bigjoolsI was going to redirect email to launchpad-reviewers00:05
wgrantThis is already super-messy, and the complex team structure doesn't make it any less messy00:05
bigjoolsexactly wgrant00:05
wgrantMP notifications are fundamentally broken00:05
wgrantIt's always going to be messy until they're fixed00:05
bigjoolsI don't understand the arbitrary refusal to set the contact address to be the same as another team00:06
bigjoolsthat's bizarre00:06
wgrantIt's not arbitrary00:06
wgrantIt's less than ideal00:06
wgrantBut it's certainly not arbitrary00:06
bigjoolsthe error on screen should say why instead of appearing to be arbitrary00:06
wgrantLaunchpad doesn't just use it as a contact address00:06
wgrantIt's also used to look up the team00:06
wgrantMostly by Soyuz00:06
* bigjools sets contact address to devnull@example.com00:07
wgrantGood luck confirming that :)00:07
bigjoolsbollocks, it ignored it!00:08
bigjoolsthis is fucking infuriating :/00:08
lifelesswgrant: I agree that things could be better, but given the *current code base* I think the simplest way to manage it is leaf teams for things that do stuff, and organisational teams put into those.00:08
lifelesswgrant: arguing that its broken just aknowledges that it is.00:08
bigjoolsis this recommended structure documented anywhere on the lp wiki?00:08
lifelesswgrant: the other reason to do it the same way as LP is that it would be the same way as LP00:09
lifelessbigjools: https://dev.launchpad.net/CreatingNewProjects00:09
wgrantlifeless: Sure, but that requires renaming branches everywhere00:09
wgrantlifeless: So it's far less destructive to set a contact address00:09
lifelesswgrant: ah, that is a cost to moving, yes.00:09
lifelesswhich will hurt.00:09
bigjoolslifeless: not sure that page helps00:10
lifelessbigjools: I don't think we captured the how-we-got-there anywhere other than the list archives :( - sorry!00:10
lifelessI have to run, shopping trip to do00:10
bigjoolslifeless: :(00:10
lifelessback soon, good luck solving your headache.00:11
lifelessmy 2c before I go00:11
lifeless - make a new team -reviewers and copy the direct member list from -maintainers to -reviewers00:11
lifelessmake it the reviewer00:11
lifelessmake it owned by -maintainers00:12
lifelessand refactor to remove duplication post-release.00:12
* bigjools made launchpad-reviewers the reviewer00:17
wgrantMy inbox thanks you00:18
bigjoolslp-reviewers was a member of the old team anyway00:19
wgrantBut now I'll get less mail :)00:19
bigjoolsthat was the idea :)00:19
bigjoolsalso, when I change the status of an MP to approved it'd be really nice if it either warned about new versions not in the page or just approved them anyway00:22
lifelesszomg its a statik00:45
wgrant2012-09-25 22:55:56 INFO    Outage complete. 0:00:01.81845500:53
wgrantgetting there00:53
wgrant(from the latest staging update)00:53
StevenKWhat, too long?00:54
wgrantNo reason for it to take more than a couple of hundred milliseconds00:55
lifelessStevenK: yeah00:56
wgrantsecurity.py is most of the remaining time, I believe00:57
wgrantStevenK: https://code.launchpad.net/~wgrant/launchpad/bug-1056506/+merge/12635800:59
=== matsubara is now known as matsubara-afk
StevenKwgrant: r=me01:01
wgrantStevenK: Thanks01:01
StevenKwgrant: I heartily approve of lines 61-6901:02
lifelesslibreoffice starts -fast- on SSD01:06
wgrant2012-09-26 01:17:38 INFO    Resetting permissions.01:18
wgrantSecurity took 0.78046512603801:18
wgrant2012-09-26 01:17:38 INFO    Enabling access to launchpad_dev_master.01:18
wgrant2012-09-26 01:17:38 INFO    Outage complete. 0:00:00.88874801:18
wgrantSo security.py is indeed probably most of it01:18
lifeless      327 /    0  DistroSeries:EntryResource:getBuildRecords02:12
lifeless      321 /    0  Person:+specworkload02:12
StevenKWhy do we get a PQM success mail at 11:59 and then a failure at 12:04?02:22
wgrantStevenK: There are lots of mail delays02:25
wgrantStevenK: So buildbot-poll sends the email02:25
wgrantIt gets stuck in a queue02:25
wgrant5/10 minutes later, buildbot-poll sees there's still stuff to merge02:26
wgrantSends the email02:26
wgrantMail queue unclogs, PQM processes both02:26
StevenKI look forward to when we can kill buildbot-poll02:26
* StevenK stabs lxc02:42
StevenKwgrant: LXC is impenetrable.03:35
wgrantStevenK: Howso?03:35
StevenKI don't understand it at all!03:36
wgrantI know parts of it all too intimately.03:36
wgrantWhat is it doing to you?03:36
StevenKI destroyed the ec2 instance in disgust before lunch, but the lxc-setup-build script in examples was hanging03:36
wgrantStevenK: Did you have a base container set up properly?03:37
wgrantAnd where was it hanging?03:37
StevenKlxc-wait -n or so, I think03:38
StevenKAnd base container?03:38
StevenKSee previous statement about impenetrability.03:38
wgrantlp-setup-lxc-build relies on there being a base container already03:38
wgrantIt builds LP inside a container03:38
wgrantYou can use lpsetup to create this container03:39
lifelessI wonder if we need more 'this is how its all put together' docs for this03:41
wgrantParticularly since the in-DC setup is different03:44
wgrantI suspect we'll be able to pull together docs from me watching StevenK trip over everything :)03:44
StevenKOh. Yay?03:45
wgrantStevenK: So, do you understand the high-level fundamentals of how parallel testing is implemented for LP today?03:46
wgrantI suspect few outside lifeless, I, Yellow, and some ops, do03:46
wgrantBut I'm not sure how much you've picked up on03:47
wgrantAs it's all a bit complicated03:47
StevenKwgrant: We create 20 containers run a portion of the tests in them, and testr interleaves it all together into a useful subunit stream03:47
wgrantStevenK: Right03:47
wgrantNow, the 20 containers aren't all created from scratch each time03:48
wgrantThey're ephemeral (created by the aptly named lxc-start-ephemeral) as clones of a base container03:48
wgrantThis base container has LP code+deps+DB etc in it.03:49
wgrantAnd it's persistent03:49
wgrantlxc-start-ephemeral uses aufs or overlayfs to create effectively COW snapshots of the base container's tree03:49
wgrantThen bring up a new throwaway container on that03:49
wgrantYou can see the buildbot config in lp:~launchpad/lpbuildbot/public03:50
wgrantWith a testr.conf that calls IIRC lp-setup-lxc-test to run bin/test in an ephemeral container03:50
StevenKYes, it calls /usr/local/bin/lp-setup-lxc-test03:51
wgranttestr just calls the command there to run the tests, so testr doesn't know about LXC at all03:51
wgrantAs far as it's concerned it's just running 20 parallel testrunners locally03:51
StevenKI was guessing I needed lp-setup-lxc-build first, but like I said, that wasn't filled with happiness03:51
wgrantRight, lp-setup-lxc-build is run as one of the earlier phases of each full testsuite run03:52
wgrantIt basically just runs 'make build schema'03:52
wgrantSo that's done in the base container, rather than separately in each cloned one03:52
wgrantBut it relies on there being a container already03:52
wgrantlpsetup init-lxc can probably create it for you03:53
wgrantIn the DC we do it manually03:53
wgrantThere's nothing really special about the container created by lpsetup03:54
wgrantYou can get something roughly equivalent by running rocketfuel-setup in a normal container from lxc-create03:55
StevenKwgrant: I'll bring up an ec2 instance in a bit, working on failing tests from my storm branch.03:55
wgrantThere were failures? :(03:55
wgrantQuery issues, or just things expecting SQLObjectResultSets?03:56
StevenKYeah, there was a callsite of handleOptional... hiding in registry/sourcepackage.py03:56
wgrantOh right03:56
wgrantThat one03:56
StevenKWhich I'd like to destroy03:56
wgrantI tried once upon a time to make the build API sensible03:56
wgrantBut didn't quite get all the way03:56
StevenKwgrant: Can I kill _get_ubuntu in that class?03:57
wgrantStevenK: That comment is prehistoric and no longer correct03:58
StevenKSo I can kill it?03:58
wgrantYou should be able to use an inline ILaunchpadCelebrities lookup instead03:58
StevenKI was going to, yeah03:58
wgrantCheck the history on .packaging03:59
wgrantI think the Ubuntu specialcase can probably be removed03:59
wgrantIt's the only callsite of priorReleasedSeries03:59
wgrantAnd seems pretty pointless03:59
wgrantBecause Ubuntu should have previous_series set properly everywhere04:00
wgrantEven in sampledata04:03
wgrantSo I'd try removing that specialcase and all the priorReleasedSeries definitions and seeing if the test suite complains04:03
StevenKThat gets this branch to -6004:03
wgrantYou removed it from DistroSeriesSet and the interfaces too?04:04
StevenKHold on, removed what?04:04
wgrantIts sole callsite is the Ubuntu specialcase in SourcePackage.packaging04:05
wgrantAnd I can't see any value in that special case.04:05
wgrantYou might be able to find something04:05
wgrantBut I really can't04:05
StevenKRight, it's terrible04:06
wgrantSince Ubuntu releases are serial, the first priorReleasedSeries is always going to be previous_series04:07
wgrantExcept when we're on a non-released non-development series04:07
wgrantLike q-series or r-series04:07
wgrantTheir first priorReleasedSeries is the current release04:07
wgrantnot quantal04:07
wgrantBut that provides no value, since with the special case removed it will fall back to quantal which will fall back to precise04:08
wgrantSo yeah04:08
wgrantDelete the special case04:08
wgrantThen probably lp-land and see if buildbot complains04:08
wgrantIt probably won't, and if it does then the fix is easy and meh.04:08
StevenK% bzr damage04:08
StevenKUsing submit branch /home/steven/launchpad/lp-branches/devel04:08
StevenKHealing: 134 lines of code04:08
wgrantYou got both priorReleasedSeries methods and their respective interfaces?04:08
StevenKAnd the *enormous* test04:09
StevenK% bzr grep priorReleasedSeries | wc -l04:10
wgrantis very relevant04:10
wgrantStevenK: The addendum there doesn't make much sense to me04:12
wgrantIf there are many unreleased series, then it will just recurse back04:12
wgrantIt's not necessary to jump straight to the last released one, AFAICT04:13
StevenKwgrant: Do you want to eyeball a diff before I commit, push and lp-land?04:13
wgrantStevenK: If you push I'll check the MP04:13
wgrantbigjools: I'm guessing that's all a blur and you don't remember the context around that MP any more?04:14
StevenKwgrant: Pushed04:15
bigjoolsnot looked, otp04:15
StevenKbigjools: Not still dealing with broker from hell?04:17
bigjoolsno thankfully04:17
wgrantStevenK: The amount of red in the incremental diff pleases my eyes.04:18
wgrantStevenK: Can you run -t packaging or so?04:21
wgrantJust in case there's something obvious04:21
StevenKwgrant: Total: 109 tests, 0 failures, 0 errors in 6 minutes 27.422 seconds.04:28
StevenKTotal: 109 tests, 0 failures, 0 errors in 6 minutes 27.422 seconds.04:28
StevenKBLEH, this double paste is getting annoying04:28
wgrantStevenK: Sounds good04:32
StevenKLet me lp-land it04:32
wgrantStevenK: Indeed04:35
wallyworldcody-somerville: you around?04:38
* StevenK stabs init-lxc for wanting lots of information04:42
wgrantwallyworld: lazr.restful already has infrastructure for detecting canonical_url changds04:43
wgrantThough I'm not sure why it's not working here04:43
wallyworldthis is lazr.restfulclient04:43
wallyworldand the code path does not detect it04:43
wgrantwallyworld: Sure, but lazr.restful detects a change and sends a 301 with Location04:44
wallyworldclearly not :-)04:44
StevenKBlink. It debootstraps?04:44
wallyworldthis is a fairly old bug too04:44
wgrantWell, yeah, it doesn't work in this case04:44
wgrantBut I'd probably investigate why it doesn't work04:44
StevenKI'm going to be unhappy if this requires an LP account04:44
wgrantRather than adding that somewhat terrible model hack04:45
wgrantStevenK: Why wouldn't it debootstrap?04:45
wgrantIt needs to create a fresh installation for the container04:45
StevenKwgrant: Wasn't expecting it is all04:45
wgrantIt should be OK with grabbing the branches from LP anonymously, though04:45
StevenKYou'd think04:45
StevenKI've fed it nonense, so we'll see04:46
wgrantwallyworld: See lazr.restful._resource.EntryManipulatingResource.applyChanges() for the code which does it on attribute changes04:47
wgrantBut I guess not on method calls04:47
lifelesswe could run an image rather than debootstrap04:47
wallyworldok, will look. i didn't realise lazr.restful detected url changes04:47
lifelessbut LXC isn't [currently] run that way04:47
StevenKlpsetup needs a bug04:48
StevenKIt pulls from archive no matter what, it should be configurable04:48
wgrantwallyworld: I'm not quite sure where method calls are done, but I know it automatically calls ObjectModifiedEvent, so there must be something there04:48
wgrantStevenK: Ah yeah04:48
wgrantStevenK: I manually ran lxc-create with MIRROR=whatever beforehand04:49
lifelesswgrant: have you had time to look at the ghosts?04:49
StevenKSadly, it's going faster than the ec2 mirror04:49
wgrantlxc-create stores a cache keyed on (distroseries, archtag)04:49
wallyworldwgrant: by "it", you mean lazr.restful i assume04:49
StevenKSo I can't complain much04:49
wgrantwallyworld: Right, sorry04:49
wgrantlifeless: Not yet, as it doesn't seem overly urgent. If you want me to I'll get onto it nowish, though04:49
wgrantI've been fixing bugs to make bigjools shut up :P04:50
StevenKThat fix should be deployed soon04:50
wgrantIndeed :)04:50
StevenKwgrant: Right, so init-lxc just finished04:51
lifelesswgrant: I'd like to know its good this week04:51
wgrantStevenK: Did it branch LP stuff into it and stuff?04:51
lifelesswgrant: as config changes when they are needed, are generally needed now04:51
wgrantlifeless: Yeah, true04:51
wgrantAnd we have no ops next week04:51
StevenKwgrant: I don't think so04:51
StevenKwgrant: How can I check?04:51
lifelesswgrant: exactly04:51
wgrantStevenK: ls ~04:52
lifelesswgrant: does the lp-setup-lxc stuff bind-mount ~?04:52
lifelessor does it do a fresh copy of everything ?04:52
StevenKroot@ip-10-78-117-158:~# ls -lh ~04:52
StevenKtotal 004:52
wgrantlifeless: it bind mounts ~04:52
wgrantStevenK: Ah right04:52
lifelessthats root though :P04:53
wgrant    init-lxc            Create an LXC container suitable for later installing04:53
wgrant                        a Launchpad development environment.04:53
wgrant    init-repo           Get the Launchpad code and dependent source code.04:53
wgrant    install-lxc         Completely sets up an LXC environment with Launchpad04:53
wgrant                        using the sandbox development model.04:53
StevenKOh, bleh04:53
lifelessStevenK: tell me you ran that as root ?04:53
StevenKlifeless: Which as root?04:53
lifelessStevenK: is this on your machine ?04:53
wgrantFrom install-lxc:04:53
wgrant    This is basically the same as running `init-lxc` and then, inside the \04:53
wgrant    newly created container, `lp-setup init-repo`, `lp-setup update`, and04:53
wgrant    `make install`.04:53
StevenKlifeless: Note the hostname04:54
lifelessStevenK: doesn't tell me anything04:54
StevenKIt's an m2.xlarge ec2 instance04:54
lifelessStevenK: thanks04:54
StevenKm2.2xlarge, rather04:54
StevenK34GiB of RAM04:54
wgrantStevenK: So I'd start the container, SSH in, and run lp-setup init-repo, lp-setup update, make install04:54
StevenKlxc-start lptests ?04:55
wgrantsudo lxc-start -n lptests -d04:55
wgrantWithout -d it'll start in the foreground and give you local console04:55
wgrantWhich is usually undesirable04:55
wgrantYou can then use lp-lxc-ip (or something close to that) to determine the IP address to give to ssh04:56
StevenKI thought it was ubuntu/ubuntu04:56
wgrantIt should be the username and password of whatever account you ran it as, usually04:56
wgrantBut otherwise ubuntu/ubuntu04:56
wgrantOr root/04:56
StevenKroot/ doesn't work either04:57
wgrantIt's possible that running it as a passwordless root account has confused it04:57
wgrantCheck /var/lib/lxc/lptests/rootfs/etc/shadow04:57
StevenKDisabled root, it has an jenkins account04:59
StevenKBut the password isn't jenkins or test or ubuntu or test04:59
wgrantAh, if you ran it as jenkins then it'll have copied your host's jenkins password04:59
StevenKThat I know04:59
wgrantI thought you ran it as root, given the PS1 earlier05:00
StevenKIt insisted I run it with -u05:00
StevenKlp-setup init-repo: error: No bzr launchpad-login set.05:00
wgrantvim is your friend05:01
wgrantI suspect05:01
* StevenK adds --use-http05:01
wgrantOr that05:01
StevenKHm, the amount of progress printed is staggering05:03
wgrantIn which direction?05:04
StevenKDo you want to proceed? [y/N] y05:04
StevenK  05:04
StevenKRight, it is doing something, it just isn't telling me05:05
* wgrant watches 2500 gpg prompts fly past05:08
wgrantI should probably have disabled signing before running mkghosts05:09
StevenK20 failures05:10
StevenK6 errors05:10
StevenKSo, I love05:10
StevenKlose, even05:10
lifelesswgrant: wheee :P05:11
wgrantStevenK: Hm, what05:11
wgrantYou lost05:11
wgrantBut how05:11
wgrantThose errors don't make much sense05:11
wgrantIt looks like the DSP is ending up None05:11
StevenKBrokenImplementation: An object has failed to implement interface <InterfaceClass lp.answers.interfaces.questiontarget.IQuestionTarget>05:12
wgrantAh, no05:12
wgrantIt's just answer_contacts that's broken, it seems05:12
wgrantBy "The answer_contacts attribute was not provided" it means "AttributeError: 'NoneType' object has no attribute 'using'"05:13
StevenKHow did I manage to screw that up?05:13
wgrantOh heh05:13
wgrant     @property05:13
wgrant-    def _store(self):05:13
wgrant-        return Store.of(self.sourcepackagename)05:13
wgrantThat would do it05:13
wgrantin sourcepackage.py05:13
StevenKOh, I thought that was unused!05:14
StevenKI even bzr grepped for it05:14
wgrantQuestionTargetMixin uses it05:14
wgrantAnyway, this is why we have 35 minute buildbot :)05:15
StevenKPutting together a testfix05:15
wgrantIt looks like the only failure, but you would do well to wait until this run finishes and then run all the failures05:15
StevenKIt's going to require a branch anyway, but yeah05:15
StevenKIt just finished05:17
* StevenK grabs the stream 05:17
wgranthttp://lpbuildbot.canonical.com/builders/lucid_lp_lxc/builds/39/steps/shell_9/logs/subunit/text should be what you want, yeah05:17
StevenKI was already wgetting when I said I was grabbing it05:18
StevenKAnd I'm now waiting on testr run --failing05:18
StevenK(and bzr on said ec2 ...)05:18
StevenKjenkins@lptests:~$ du -sh launchpad/lp-branches/.bzr05:19
StevenKRan 84 (-22077) tests in 131.750s (-1768.113s)05:20
StevenKPASSED (id=1, skips=6)05:20
StevenKOMG, it finished05:21
wgrantYour ec2 instance?05:21
StevenKlp-setup update depends on CWD? :-(05:22
wgrantas do lp-setup-lxc-test and lp-setup-lxc-build05:22
wgrantIt makes a fair bit of sense05:22
StevenKjenkins@lptests:~/launchpad/lp-branches/devel$ lp-setup update05:23
StevenKlp-setup update: error: the target dir /home/jenkins/launchpad/lp-branches/devel is not a valid Launchpad tree.05:23
StevenKlp-setup update: error: the target dir /home/jenkins/launchpad/lp-branches/devel is not a valid Launchpad tree.05:24
wgrantNot seen that before05:24
wgrantLuckily the code is quite sensible05:24
StevenKOh, it's in sandbox, for ... some reason05:24
StevenKTestfix landed05:25
wgrantStevenK: Right, it uses lightweight checkouts05:27
wgrantdevel is a treeless branch05:27
wgrantsandbox is a lightweight checkout05:27
StevenKAnd devel is empty05:27
wgrantIt has a .bzr05:27
wgrantBut that's it05:27
lifelesswgrant: StevenK: more low hanging fruit: https://bugs.launchpad.net/loggerhead/+bug/86930905:36
_mup_Bug #869309: If folders in repository have unicode names, sometimes I get KeyError <navigation> <oops> <trivial> <unicode> <loggerhead:Triaged> < https://launchpad.net/bugs/869309 >05:36
lifelessbrought to my attendion by the new user that just assigned it to themselves ><05:36
wgrantlifeless: But there's no low-hanging fruit :)05:37
lifelesswgrant: it has a patch attached.05:37
lifelesswgrant: doesn't get much lower :)05:37
StevenKwgrant: Right, that's all done on my ec2 slave05:42
=== almaisan-away is now known as al-maisan
wallyworldis it just me or does anyone else's screen dimming fail to turn the screen back when the mouse is moved? the pointer shows up but the screen stays black :-(05:45
lifelesswallyworld: that could be a dropped vsync for unity05:47
lifelesswallyworld: it looks like a dimming problem, but it isn't05:47
wallyworldit doesn't happen if i turn off "screen dimming after 10 minutes"05:48
wallyworldso i just got back from making a coffee and boom :-(05:48
wallyworldcause i forgot to turn it off again05:48
lifelesschat to RAOF05:49
lifelessin #ubuntu-dekstop05:49
lifelessonce they finish figuring out where to put the privacy setting05:49
wallyworldok, i'll search bugs first05:50
StevenKwgrant: So, next step?05:51
wgrantStevenK: Do stuff05:53
wgrantStevenK: In sandbox on the host, run lp-setup-lxc-build05:53
wgrantSee if it works05:53
* StevenK stabs it for prompting for a sudo password every 2 seconds05:57
StevenKwgrant: In a word, no06:04
wgrantbuildbot loves you, though06:04
wgrantStevenK: What broke?06:05
wgrant"Everything" is an acceptable answer.06:05
StevenKThe script06:06
StevenKwgrant: http://pastebin.ubuntu.com/1227964/06:07
wgrantThis command has to be run as root06:07
wgrantsudo it?06:07
wgrantYou did06:07
wgrantStevenK: Why ~/lp-setup-lxc-build rather than the one installed in PATH?06:08
wgrantDid you hack it06:08
StevenKThere is none in PATH06:08
wgrantlp-setup should have come with one06:08
wgrantI thought06:08
StevenKIt's in /usr/share/pyshared/lpsetup/templates/lp-setup-lxc-build06:09
StevenKAnd I got sick of that long path so I copied it and +x'd it06:09
wgrantContainer {lxc_name} does not exist06:09
wgrantSounds like you didn't replace the bits of the template?06:09
StevenKOh, I need to sed over it?06:09
wgrantIt takes {lxcip}, {lxc_name} and {ssh_key_path} variables06:10
wgrantThe last is just a local SSH key to use to connect to the containers06:10
StevenKSo I need to fix it, it certainly doesn't use $1 and so on?06:11
wgrantlxcip shouldn't be an arg, but the other two arguably should be06:11
StevenKwgrant: So the name is lptests or something that will based off it, like current-test-0 or something?06:27
wgrantStevenK: lptests is the base, the ephemerals will be something like lptests-temp-wefwhjefuwehfuiweghfuiw06:28
StevenKwgrant: Sorry, I mean what should {lxc_name} be?06:29
wgrantStevenK: lptests06:29
StevenKwgrant: So lxc-stop on lptests failed with Operation not permitted, but it seemed to work06:35
lifelesslp:~lifeless/python-oops-datedir-repo/bug-971255 <- another critical for you, MP coming up06:36
wgrantStevenK: You need to sudo lxc-* operations, usually06:36
StevenKExcept the whole script is running as root06:37
wgrantThat's a little odd06:38
StevenKHm, I think I should done this as an EBS backed instance06:39
StevenKwgrant: https://qastaging.launchpad.net/builders/palmer/+history?build_text=libc&build_state=all consistenly times out :-(06:40
lifelesswgrant: StevenK: or wallyworld: ^06:40
StevenKNo diff06:41
StevenKNo review06:41
lifelessStevenK: there is a diff06:41
lifelessStevenK: thanks to the magic of ajax06:41
StevenKLooks good06:42
wgrantStevenK: Indeed, but we're OK as long as the stuff that worked before hasn't regressed.06:42
wgrantWe should probably just drop name filtering from that page eventually06:43
* StevenK destroys this instance06:44
wgrantStevenK: I'd recommend against using EBS, as that would let you work around setup awkwardness rather than fixing it properly06:45
StevenKwgrant: I'm trying to avoid the two hour wait lp-setup install-lxc brings06:45
lifelesswgrant: StevenK: wallyworld: https://code.launchpad.net/~lifeless/python-oops-datedir-repo/bug-1003627/+merge/126392 - with this I'll do a release, and we'll have that critical fix-released.06:59
StevenKIs the prune bug already solved?07:00
lifelessthe pruning too fast one is07:00
lifelessthis is pruning-the-right-stuff-is-impossible.07:01
lifelesswe have to physically partition the datedir repo for things in different project groups07:01
lifelessthis will fix it07:01
StevenKDidn't wgrant write a patch for the prune reference bug?07:01
lifelesshe did: 19:00 < lifeless> the pruning too fast one is07:02
lifelessthis is a different prune bug07:02
StevenKlifeless: r=me07:03
lifelesswgrant: isn't bug 1050722 FR ?07:17
_mup_Bug #1050722: prune only searches for OOPS references up to prune_until <Python OOPS Date-dir repository:Fix Released by wgrant> <python-oops-tools:In Progress by wgrant> < https://launchpad.net/bugs/1050722 >07:17
wgrantlifeless: Ah, indeed07:18
wgrantClosed, thanks.07:18
lifelessStevenK: https://code.launchpad.net/~lifeless/python-oops-tools/bug-1003627/+merge/126401 if you would07:33
lifelessStevenK: I've set a commit message, so if you can toggle the MP to approved (if you do approve) that would rock. Thanks!07:34
* lifeless waves bye-for-a-bit07:34
StevenKDo not understand why prune is duplicated07:35
adeuringgood morning07:48
lifelessStevenK: the CLI is the only duplicate07:56
lifelessStevenK: oops-tools prunes the DB07:56
lifelessStevenK: datedir-repo prunes disk07:56
lifelessStevenK: and the API client is in just one place07:57
StevenKYes, but you're duplicating logic between them07:57
lifelessStevenK: the CLI only07:57
wgrantThere is some code that we should dedupe further07:59
wgrantBut it's not completely terrible atm07:59
wgrantand it can't all be generalised07:59
lifelesswriting a disk-dir implementation of the django orm terrifies me08:04
lifelessStevenK: thanks!08:04
lifelessStevenK: wgrant: I'll leave you(purple) to get the prune cronjobs raionalised at a convenient08:05
wgrantOf course :)08:06
lifelesswgrant: thanks08:10
mvohello, it seems like exports for http://bazaar.launchpad.net/~ubuntu-core-dev/ddtp-ubuntu/ddtp-quantal/changes  stopped at Aug 16th for some reason, would it be possible to check if something is stuck/stale (lockfile etc)? it would be nice to get a updated package description export for the ubuntu beta208:12
czajkowskiwgrant: StevenK wallyworld_ one for you guys to kick or invstigate ?08:12
* wallyworld_ runs away (my packing foo is zero)08:13
wallyworld_packaging even08:13
dpmwallyworld_, there's no escape. It's not about packaging, it's about bzr branches ;)08:14
wgrantwallyworld_: It's actually translations08:14
wgrantWhich none of us have any clue on08:14
wgrantBut let's see08:14
wgrantThat date is a little suspicious08:14
wgrantI wonder if the import was interrupted by a DC move and left a lock around08:14
wgrantSince ddtp exports take forever08:14
wallyworld_ah, is that when the move happened08:15
wgrantAugust 18, but yeah08:15
wgrantRather close08:15
wgrant2012-09-26 05:11:44 ERROR   Failure in ddtp-ubuntu/quantal: LockContention(Could not acquire lock "(local)": )08:15
* mvo has no clue either but hopes it just a "rm export-ddtp-quantal.lock"08:16
wgrantmvo: bzr break-lock lp:~ubuntu-core-dev/ddtp-ubuntu/ddtp-quantal08:16
wgrantshould do it08:16
wgrantHopefully it'll say there's a lock there from taotie on Aug 1808:16
wgrantIt wasn't killed by the move, but exports were severely broken due to fallout from the move a day after08:16
mvowgrant: hm, that did just return, no message no nothing?08:17
wgrantLet's see08:18
wgrantjam, jelmer, mgz: ^^ any ideas on the above?08:23
czajkowskiwallyworld_: StevenK could one of you please help me out on https://answers.launchpad.net/launchpad/+question/20946508:28
wallyworld_czajkowski: i think if they use one of the supported bug trackers eg bugzilla etc then lp will import bug data but it will still be available in the external system too08:29
wallyworld_but i'm not 100% sure what happens the other way08:30
jamwgrant: http://bazaar.launchpad.net/~ubuntu-core-dev/ddtp-ubuntu/ddtp-quantal/.bzr/repository/lock/ and http://bazaar.launchpad.net/~ubuntu-core-dev/ddtp-ubuntu/ddtp-quantal/.bzr/branch/lock/08:31
jamshow that no lock is currently held.08:31
czajkowskiwallyworld_: ok thanks :)08:31
wgrantjam: Exactly08:31
jamI don't really knowt about the exception you saw.08:31
wgrantczajkowski: They won't be able to search for or report new bugs on Launchpad, but all data will still be there if they choose to reenable it08:31
jamIt is possible to be a semi-genuine contention, if you are re-opening a branch in a script, and now one side is holding a lock, and the other is trying to do something.08:31
wgrantjam: It's only been happening on that branch since the move chaos08:32
wgrantWithout code changes08:32
wgrantSo it seems unlikely to be legitimate.08:32
wgrantI don't see howit can be "(local)" but without a URL08:32
jamwell, we did upgrade bzr in the middle term08:32
wgrantIt worked on Aug 16, we moved to a new server with significant drama on Aug 18, then it hung a bit until Aug 30, and from Aug 31 it's been giving that error08:34
wgrantWas there a bzr upgrade in those two weeks? Possibly...08:34
cjwatsontumbleweed: Oh, are you still looking for a db review for your edit-packagesets branch?  https://dev.launchpad.net/PolicyAndProcess/DatabaseSchemaChangesProcess has instructions on whom to seek db reviews from08:42
tumbleweedcjwatson: aha, I never looked there. Thanks08:54
mvoczajkowski, wgrant: sorry, I was in a meeting now, should I rather file a bugreport about the export issue instead of asking on irc? especially if its not as trivial as I had hoped?09:52
czajkowskimvo: probablly best that way the maintenance team can look at it09:59
StevenKWhich export issue?10:00
mvoStevenK: lp:~ubuntu-core-dev/ddtp-ubuntu/ddtp-quantal is not getting translation updates anymore since ~1month10:01
mvoI filed bug https://bugs.launchpad.net/launchpad/+bug/1056752  now10:09
_mup_Bug #1056752: Translations export not working for lp:~ubuntu-core-dev/ddtp-ubuntu/ddtp-quantal <Launchpad itself:New> < https://launchpad.net/bugs/1056752 >10:09
rick_h_thanks for the qa wgrant10:50
* mpt wonders what "div.first{*margin-right:4%;_margin-right:1.3%}" is supposed to do11:23
mgzscrew with IE users for fun?11:24
mptNeither *margin-right nor _margin-right is a valid property, as far as I know11:25
mgzthe star property hack targets IE 7 and less, the underscore hack targets IE 6 and less11:29
rick_h_css hacks ftw11:29
mptoh I see11:29
mgzso, it's deliberatly invalid css to get older buggy browsers that misinterpret it as valid to apply specific rules11:30
mgzbut... that's pretty bad, setting percent margins is likely to just break things anyway11:30
mptI was just trying to figure out why bug reports had become unreadable, and it was because my custom Launchpad CSS had stopped working11:30
mgzand having different values for different browsers then makes working out why something looks funny even more entertaining11:31
rick_h_ah, well % css rules are generally a good thing as they keep things in scale for different browser sizes11:31
rick_h_we're not in the land of pixel perfect and such11:31
rick_h_curious what custom CSS you run with.11:31
mptyay, fixed11:32
mptrick_h_, http://paste.ubuntu.com/1228291/11:33
rick_h_mpt: cool11:33
rick_h_while I like smaller fonts I've often thought a lot of users must find it small11:34
=== gary_poster|away is now known as gary_poster
rick_h_jcsackett: ping when you get a sec12:35
=== matsubara-afk is now known as matsubara
=== abentley changed the topic of #launchpad-dev to: http://dev.launchpad.net/ | On call reviewer: abentley | Firefighting: - | Critical bugs: ∞
=== slank` is now known as slank
=== slank is now known as Guest84818
rick_h_abentley: a review for you if you get a chance please https://code.launchpad.net/~rharding/launchpad/info_type_events/+merge/12631714:10
rick_h_jcsackett: would appreciate your look as well if you have the bandwidth. ^^14:10
rick_h_I think this will address the manual usage and we'll see I guess when I follow up and integrate the change into place.14:11
abentleyrick_h_: What happens if an item is private and a feature on the page is in beta?  That should show two banners, but you've made it a singleton, IIUC.14:15
rick_h_abentley: the beta code is in betabanner.js and has it's own singleton14:16
abentleyrick_h_: cool.14:16
rick_h_that's just moved code, from the bottom to the top of the file so it's how it currently works14:16
abentleyrick_h_: "@namespace lp.app.banner" reminds me that we haven't reached consensus about whether to separate namespaces from module names.14:17
abentleyrick_h_: But you didn't change that, so that's alright.14:18
rick_h_abentley: yea, at some point I want to put together some notes/examples and see14:18
rick_h_I looked at a rename here, but it hit some 50 files and I didn't want to get into it with this14:19
rick_h_because we've started a lp.ui namespace and this would be good as lp.ui.banner.PrivacyBanner, lp.ui.banner.BetaBanner, etc14:19
rick_h_but for another day14:19
jcsackettrick_h_: ping. i lost history in irc, so the details as to why you needed me to ping you are lost to me. what's up?14:22
abentleyrick_h_: Why attach the _singleton_privacy_banner to window?14:22
rick_h_abentley: to be explicit. That's what's happening itself, but making it explicit makes sure in all cases you're using a global instance14:23
rick_h_jcsackett: sorry, ping'd earlier because I didn't get how banner rendering worked with the .pt bits14:23
danilosderyck, hi, I've filed a few bugs related to specification privacy: bug 1056881, bug 1056891 and bug 1056893 - sorry if I made any duplicates, but I didn't find them14:23
_mup_Bug #1056881: Specifications privacy: subscribers can't see private blueprints <Launchpad itself:New> < https://launchpad.net/bugs/1056881 >14:23
_mup_Bug #1056891: Specification privacy: upcoming work won't load for anyone if there's any private data on it <Launchpad itself:New> < https://launchpad.net/bugs/1056891 >14:23
_mup_Bug #1056893: Specification privacy: no "private content" header on +upcomingwork page with private data <Launchpad itself:New> < https://launchpad.net/bugs/1056893 >14:23
rick_h_jcsackett: but basically just ping'd on the review since you looked at it yesterday with feedback14:23
rick_h_jcsackett: feel free to ignore14:24
jcsackettrick_h_: so all is well now? i'm happy to answer questions whenever. it's not the clearest of code, despite my best efforts at the time. :-P14:24
rick_h_jcsackett: well, I broke off rendering into a second step. Just events for now14:25
jcsackettrick_h_: dig.14:25
rick_h_so might bug you later on, but for now I"m cool14:25
jcsackettrick_h_: righto.14:25
abentleyrick_h_: I don't understand the value of that explicitness.  Because there's no encapsulation, there's no longer a guarantee that it is a singleton.14:26
rick_h_abentley: window is a browser enforced singleton global space. It's to make sure you know where that's at and prevents people from looking for var definitions and such14:26
rick_h_for instance, in tests, I know I need to reset that window.var14:26
deryckdanilos, thanks for filing the bugs.14:29
abentleyrick_h_: Window is just a global.  It doesn't ensure there's only one instance of PrivacyBanner.  If I can set __singleton_privacy_banner to null, then getPrivacyBanner will create a second instance.14:29
rick_h_abentley: right, but the module is only loaded once. So it's set to null one on module load. And then never again14:29
danilosderyck, yw, thanks for getting this in, Linaro wants to use it asap :)14:29
rick_h_except in tests where I do that to do a proper teardown14:30
abentleyrick_h_: Sorry, forgot I was reviewing before the call.15:06
abentleyrick_h_: It doesn't seem very decoupled to be emitting privacy_banner:hide/privacy_banner:show.  If you know about the privacy banner, why not ns.getPrivacyBanner()._make_public ?15:12
rick_h_abentley: because in that case you need to deal with the instance15:14
rick_h_abentley: the future branch will get rid of the singleton code so you won't call anything. You'll just fire the event.15:14
rick_h_and ideally, you'll be firing the information type event, not the privacy banner event15:15
rick_h_so the calling code won't be using _make_public. What you'll have is banners and portlets listening for information_type:change/is_public/is_private events15:15
abentleyrick_h_: Okay, makes sense.15:15
abentleyrick_h_: r=me.15:24
rick_h_abentley: cool thanks for the review15:24
=== al-maisan is now known as almaisan-away
=== matsubara is now known as matsubara-lunch
jcsackettabentley: you available to look at a branch that just fixes up sampledata?16:03
abentleyjcsackett: sure.16:03
jcsackettabentley: awesome. :-) https://code.launchpad.net/~jcsackett/launchpad/remove-date_next_suggest_packaging-needs-sampledata/+merge/12647516:03
abentleyjcsackett: r=me.16:04
jcsackettabentley: thanks!16:04
robbiz^Hi guys! I've some problems installing launchpad on my laptop, with chroot and rocketful-setup16:10
robbiz^can I ask someone of you how to fix it?16:10
czajkowskirobbiz^: you might have more luck if you start to explain whats going on16:22
czajkowskior else posting to launchpad-dev mailing list16:23
czajkowskirobbiz^: https://launchpad.net/~launchpad-dev16:24
robbiz^Well my  postgresql database didn't start16:27
robbiz^the message is this16:27
robbiz^Error: The cluster is owned by user id 120 which does not exist any more16:28
robbiz^going through system folders I've no config file in /etc/postgresql/9.1/main16:29
robbiz^I have Ubuntu precise with chroot system as explained in dev.launchpad.net16:34
jcsackettsinzui: you free to talk for a bit?16:46
robbiz^Ok... Probably it's better to post on the mailing-list!16:50
sinzuijcsackett, ding dong17:18
sinzuijcsackett, I think this test should pass, but it does not. It might be the origin of the 404 because mail rationales and canonical urls are need to explain why the email as received: http://pastebin.ubuntu.com/1228850/17:22
jcsackettsinzui: awesome.17:37
=== deryck is now known as deryck[afk]
=== almaisan-away is now known as al-maisan
=== al-maisan is now known as almaisan-away
=== matsubara-lunch is now known as matsubara
=== Guest84818 is now known as slank
=== slank is now known as Guest32361
=== Ursinha is now known as Ursinha-afk
=== Ursinha-afk is now known as Ursinha
=== deryck[afk] is now known as deryck
lifelessjcsackett: 'When this is improved' ? <- approved??20:00
jcsackettlifeless: Sample data branch?21:33
jcsackettlifeless: approved, yes.21:39
lifelessjelmer: hey, still interested in packaging lptools ?21:54
jelmerhi lifeless21:55
jelmerlifeless: no, not really after wheezy21:55
lifelessjelmer: have you been using bzr builddeb etc for it, or just regular packaging ?21:58
lifelessjelmer: I see 'new upstream snapshot' in the log - can't quite tell :)21:58
jelmerlifeless: I only use bzr builddeb for packages in bzr so that must be what I was using :-)21:59
lifelessjelmer: also, I haven't used the debian bzr facilities for a couple years - whats the current stuff for pushing there...21:59
lifelessjelmer: or perhaps I can convince you to do the upload of 0.2.0 and ITA it at the same time ?21:59
jelmerthe last branch is @ http://anonscm.debian.org/bzr/bzr/collab-maint/lptools/unstable/21:59
jelmerlifeless: Sure, I'm happy to do another upload22:00
jelmerlifeless: do you mean RFA rather than ITA?22:00
lifelessyes, crossed wires22:00
lifelessI was thinking intent to abandon :)22:00
jelmerlifeless: but sure, happy to do another upload with 0.2.0 and the RFA22:01
lifelessjelmer: that would be awesome, thank you22:02
lifelesslet me file it22:02
lifelessjelmer: actually, I'm going to O it22:08
jelmerlifeless: euhrm, okay22:09
lifelessjelmer: I doubt there will be another upload before I finish here22:09
lifelessjelmer: and after that I'll have other things to worry about, so saying I'd maintain it in the interim would be an exaggeration22:09
jelmerlifeless: sounds reasonable to me22:10
lifelesshttp://www.debian.org/devel/wnpp/ : "RFA...hey will maintain it in the meantime, but perhaps not in the best possible way. In short: the package needs a new maintainer. "22:10
jelmerlifeless: I originally thought that's what you meant22:10
lifelessThe package has been "Orphaned". It needs a new maintainer as soon as possible22:10
czajkowskiwgrant: see pm22:10
jelmersince I think a RFA doesn't require changes to the package itself22:10
lifelessjelmer: I'm not sure O does either does it ?22:10
lifelessjelmer: adoption requires the change22:11
jelmerlifeless: it does, requires changing the maintainer to Debian QA team22:11
lifelesshuh, inconsistent docs ftw22:11
lifelessjelmer: so, whats easiest for you?22:11
jelmerlifeless: if you file the O bugreport I can do an upload that changes the maintainer to the QA team22:12
lifelesswill do22:12
=== gary_poster is now known as gary_poster|away
lifelessjelmer: just waiting for the ack22:16
lifelessjelmer: bug 68891622:18
_mup_Bug #688916: 75-80% CPU-Usage (starting with 1.9.0beta1) <cpu> <Mixxx:Invalid> < https://launchpad.net/bugs/688916 >22:18
lifelesswell, not that one :P22:18
=== abentley changed the topic of #launchpad-dev to: http://dev.launchpad.net/ | On call reviewer: - | Firefighting: - | Critical bugs: ∞
jelmerwe've given up on critical bugs?22:21
=== StevenK changed the topic of #launchpad-dev to: http://dev.launchpad.net/ | On call reviewer: - | Firefighting: - | Critical bugs: âˆ~300
=== StevenK changed the topic of #launchpad-dev to: http://dev.launchpad.net/ | On call reviewer: - | Firefighting: - | Critical bugs: ~300
sinzuiwgrant, https://code.launchpad.net/~sinzui/launchpad/bug-subselect-timeout/+merge/12656622:58
StevenKwallyworld__: http://theamazingios6maps.tumblr.com/23:10
sinzuijcsackett, r=me23:16
jcsackettsinzui: awesome, thansk.23:17
sinzuiStevenK, https://devpad.canonical.com/~curtis/launchpad-critical-report/launchpad-criticals.html is updated.23:19
StevenKwgrant: So, lp-setup-lxc-build actually does a build23:33
wgrantStevenK: Is the build successful?23:34
StevenKIt looks to run make schema too, it looks fine23:35
StevenKwgrant: lxc-stop: failed to stop 'lptests': Operation not permitted and ls: cannot access lptests: No such file or directory are a little concerning23:36
wgrantStevenK: You sure it's adequately root?23:37
StevenKwgrant: It's running under sudo ...23:37
StevenKsudo lxc-stop -n lptests does not complain23:38
StevenKI may have to reach for set -x or run 'id' in the script23:38
wgrantThe apparmor workaround shouldn't be necessary to get this bit working23:39
wgrantJust some stuff inside the container23:39
wgrantLike postgres23:39
wgrantAnd I think lpsetup does the apparmor disablement.23:39
StevenKApparently not for nested23:39
StevenKAccording to the README23:39
StevenKBut who reads those anyway23:39
StevenKwgrant: So the next step is to fire up lp-setup-lxc-tests?23:43
wgrantStevenK: Right, fill in the lp-setup-lxc-test template and then run something like 'lp-setup-lxc-tests -t lp.services.config' and see what happens23:45
StevenKPain, misery and suffering is my guess23:46
wgrantAnd/or great success.23:46
StevenKWell, those aren't mutally exclusive. :-)23:46
wgrantStevenK: Any luck?23:53
StevenKwgrant: Just about to run it23:53
StevenKForgot {ssh_key_path}23:54
wgrantIt's easy to do23:55
StevenKRargh, single quotes23:58

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!