/srv/irclogs.ubuntu.com/2014/04/15/#juju.txt

=== CyberJacob is now known as CyberJacob|Away
=== timrc is now known as timrc-afk
Kupo24zHey all, I've setup a flat-networking openstack cluster with juju however the networking config area seems to be missing in horizon. Nova network shows running on all compute nodes however. Any ideas?04:25
=== timrc-afk is now known as timrc
josenegronjl: ping05:09
negronjljose: hey05:15
josenegronjl: hey! he estado trabajando en el charm de Seafile que habías hecho, y ya lo tengo terminado, sólo me falta un copyright file, que vendría de tu parte05:15
joseno sé si puedas poner uno en tu branch para poder hacer un pull y encargarme de mandarlo a la Charm Store05:16
negronjljose: submit it and I'll add it05:16
josenegronjl: you mean, should I push my branch and you'll put an MP to add it?05:17
negronjljose: you can add this one: https://github.com/haiwen/seafile/blob/master/LICENCE.txt05:18
negronjljose: also, add your name to the copyright ( for the charm )05:19
josenegronjl: I just did some tweaks, you're actually the author :)05:19
negronjljose: that should suffice then, submitted for review and we can see how it goes05:19
negronjljose: add my name to the copyright as well ... push it and I'll review it05:19
negronjljose: or just MP the thing and I'll take it from there05:20
joseok, I'm finishing up the README and the icon and I'm pushing it05:20
negronjljose ... oh ... and thanks for fixing it :)05:20
joseno worries :)05:20
josenegronjl: https://code.launchpad.net/~jose/charms/precise/seafile/trunk if you want to take a look05:50
=== timrc is now known as timrc-afk
negronjljose: le haces MP ?  Asi yo hago el review y por ahi seguimos05:52
josedale05:52
jose(el bug con charmers para la inclusion en el charm store ya está abierto)05:53
negronjljose: Lo miro en un rato05:54
josegenial, gracias :)05:54
negronjljose: Thanks for the changes.  I merged the changes06:18
josenp, looking forward to seeing it on the store soon06:18
=== CyberJacob|Away is now known as CyberJacob
negronjljose: I added you to the maintainers as well.06:19
joseawesome, thank06:20
joses06:20
joseI'm planning on adding memcached and postfix support in the future, looks promising06:20
=== vladk|offline is now known as vladk
=== CyberJacob is now known as CyberJacob|Away
benonsoftwareHiya07:54
benonsoftwareI'm currently making a charm for folding@home / origami and I'm not sure if I understand the 'provides:' section in metadata.yaml07:54
=== wesleymason is now known as wes_
=== wes_ is now known as wesleymason
=== timrc-afk is now known as timrc
=== avoine1 is now known as avoine
Kupo24zHey all, anyone know how to add a fixed range in juju openstack without it being overwritten (nova.conf)?13:51
=== mmelo_ is now known as mmelo
=== avoine1 is now known as avoine
=== benrodrigue is now known as brod
timrcHrm when I deploy a local jenkins environment using a mixture of precise and trusty containers, the precise containers just hang :( 1.18.1-trusty-amd6414:42
timrcI creating an lxc container by hand with -r precise to ensure that was the problem14:43
timrcThe precise containers are forever in a "Pending" state14:43
joseI think there was a bug about that14:48
mattywanyone seen this error before? WARNING failed to write bootstrap-verify file: cannot make S3 control bucket: A conflicting conditional operation is currently in progress against this resource. Please try again.14:57
mattywduring a bootstrap14:57
josemattyw: that's an S3 error, is your bucket name unique and you have enough privileges for bucket creation?14:59
mattywjose, everything was working until I tried switching regions about 30 mins ago, I switched back now and get this problem15:00
mattywI've logged into my account, deleted all buckets15:00
joseaccording to Amazon, that's an 409 Conflict error, with code OperationAborted15:01
smarterso, I'm running juju on precise and doing a local deployment with saucy machines, and they all fail with some apparmor error: http://sprunge.us/XgPJ15:02
smarterpreviously, I was deploying with raring machines and it worked fine15:02
josemattyw: apparently bug #1183571 is related to that15:03
_mup_Bug #1183571: bootstrap fails: A conflicting conditional operation... <cmdline> <hours> <juju-core:Triaged> <https://launchpad.net/bugs/1183571>15:03
mattywjose, I changed my control bucket id and it seems to work ok15:03
joseawesome then15:03
smarterany idea?15:03
mattywjose, thanks for the link to the bug15:03
mattywanyone seen this error when trying to deploy a local charm? ERROR error uploading charm: cannot update uploaded charm in state: not okForStorage15:09
jcastrolazyPower, you're on review this week iirc?15:14
lazyPowerjcastro: yeah, its going to be a repeat of the last 2 weeks - if you run into some high prority stuff please make a note on the board for me and i'll hit it up EOD15:14
jcastroThis afternoon I'll try to hit the easy ones and at least +115:14
lazyPowerMuch appreciated my man15:15
smarteris there any way to disable apparmor in the local instances to work around http://sprunge.us/XgPJ ?16:13
smarteroh apparently, the problem was fixed in dbus: https://launchpad.net/ubuntu/saucy/+source/dbus/+changelog16:28
smarteris there any way to run apt-get update && apt-get dist-upgrade on the lxc containers spawned by juju, before anything else?16:28
marcoceppismarter: you can run juju add-machine to enlist a bunch of machines for deploying to, then run an ssh loop to run those commands16:31
smarterokay that could work, is there any way to automate that?16:33
smarterwait actually saucy already has dbus 1.6.12-0ubuntu10 so that's probably not the cause of my problem17:00
jcastromarcoceppi, reminder to put your juju logging plugin thing in github pls17:11
smartermarcoceppi: I can't even juju ssh to the lxc to disable apparmor, because ssh to lxc containers is broken...17:16
marcoceppismarter: no it's not. What version of juju are you using?17:16
marcoceppithat was fixed in 1.1817:16
smarter1.16.x17:16
smarterah17:16
smartermaybe I should try 1.18 again, but it broke other stuff iirc17:16
smartersigh17:16
lazyPowermarcoceppi: i wish juju pprint was included by default17:17
lazyPowersuch a handy plugin17:17
jcastroERROR state/api: websocket.Dial wss://10.0.3.1:17070/: x509: certificate signed by unknown authority (possibly because of "crypto/rsa: verification error" while trying to verify candidate authority certificate "juju-generated CA for environment \"local\"")17:21
jcastroanyone  see that before? Fresh deploy on LXC on trusty17:22
jcastrolazyPower, actually I'd prefer normal juju status to be not so wordy by default17:22
joseguys, what should be the process in case I want to be the maintainer of a charm?17:25
marcoceppijose: is the charm unmaintained?17:27
josemarcoceppi: no, but the maintainer hadn't pushed a fix for ~1.5 years17:28
marcoceppijose: we don't really have good criteria for what constitues an unmaintained charm, which one is it?17:28
josemarcoceppi: owncloud17:29
=== roadmr is now known as roadmr_afk
=== vladk is now known as vladk|offline
jcastrojose, I would snag it, but send nathwill a courtesy email17:38
josejcastro: will do17:38
cory_fuWhy do I occasionally get this error: http://pastebin.ubuntu.com/7256493/17:47
cory_fuSeems like it's failing on the log command17:47
cory_fuBut it only happens every once in a while17:47
smarterokay, so I have this error http://sprunge.us/XgPJ when trying to do a local deployment from a precise host to either saucy or trusty instances18:04
smarterbut raring works fine18:04
jamespagethis might be a dumbass question but why when I do juju upgrade-charm for a charm deployed from a local branch does juju not just replace whats on disk in the units with what I have locally?18:05
marcoceppiit pretty much does that, in that it zips the contents, uploads to the juju storage service, then has all the units pull the zip down, extract over your CHARM_DIR and run upgrade-charm hook18:12
marcoceppijamespage: ^18:12
jamespagemarcoceppi, hmm18:12
jamespagemarcoceppi, I'm having to use --switch to force a wholesale replacement of the charm18:14
=== CyberJacob|Away is now known as CyberJacob
=== roadmr_afk is now known as roadmr
=== BradCrittenden is now known as bac
cory_fuGetting an error when trying to commit in bzr:19:01
cory_fubzr: ERROR: Cannot lock LockDir(http://bazaar.launchpad.net/~johnsca/charms/precise/apache-allura/refactoring-with-tests/.bzr/branch/lock): Transport operation not possible: http does not support mkdir()19:01
cory_fuHow do I fix that?19:02
cory_fuI did set my launchpad-login already19:04
cory_fuAh.  bzr bind lp:... instead of just setting the push branch, fixed it19:10
cory_fuI'm now consistently getting this error: http://pastebin.ubuntu.com/7257083/19:20
cory_fuThe juju-log helper works fine, and then a few steps later in the hook suddenly fails.19:20
cory_fuHere is the hook for reference: http://bazaar.launchpad.net/~johnsca/charms/precise/apache-allura/refactoring-with-tests/view/head:/hooks/install19:20
cory_fumbruzek, can you perhaps help me with that ^?19:26
mbruzekYeah19:26
mbruzekJust dug up the url for you19:26
mbruzekcory_fu, https://docs.google.com/a/canonical.com/document/d/19JyDGyiVqFi4K66yCp3iYH-TbrZvBmZceq-LnmPbj0w/edit#19:26
cory_fuThanks19:26
mbruzekcory_fu, Do you want to get on a G+ hangout ?19:43
cory_fuSure19:43
timrcmarcoceppi, All but one of my services are starting when I bootstrap and deploy a new local environment.  Looks like juju is only able to start one machine.  The rest sit in pending forever.  Have you encountered this recently?19:46
timrcmarcoceppi, experiencing this with both 1.17.7 and 1.18.119:46
marcoceppitimrc: I have not, but I have not used local in quite a while19:46
timrcah hm19:46
marcoceppitimrc: what does all-machine.log in ~/.juju/local/log look like?19:46
timrcmarcoceppi, I see a lot of this: machine-0: 2014-04-15 19:39:05 WARNING juju.worker.instanceupdater updater.go:231 cannot get instance info for instance "": no instances found19:48
=== Ursinha is now known as Ursinha-afk
=== Ursinha-afk is now known as Ursinha
=== mmelo_ is now known as mmelo
timrcmarcoceppi, http://pastebin.ubuntu.com/7257403/ -- I removed just the bit of log that shows the one service deploying to the only machine that started20:19
timrcmarcoceppi, machine-0 and machine-1 seem to start without issue, but machine-2 and on do not... no mention of them in the log either20:20
timrcmarcoceppi, Looks like you're having some connection issues... did you get my pastebin?20:36
lazyPowerjcastro: VEM charm is promulgated21:15
lazyPowerany other high prority items in the queue that need callout? otherwise i'm going top to bottom21:15
jcastrolazyPower, those mysql ones would be nice21:16
lazyPowerjcastro: the 3 you linked earler have already been reved/acked21:16
jcastrooh awesome, <321:17
lazyPower:) Context switching all day my man. i know whats up21:17
=== mmelo_ is now known as mmelo
joselazyPower: hey, mind reviewing an MP I just did to unattended-upgrades? it's pretty simple and straightforward, just 3 lines modified21:28
josewell, 421:28
lazyPowerjose: link to MP?21:28
joselazyPower: https://code.launchpad.net/~jose/charms/precise/unattended-upgrades/add-categories-readme-markdown/+merge/21596621:28
lazyPowerjose: nix the $'s from the commands in the markdown21:30
lazyPoweruse CODE output, which you have, and i'll ack this21:30
joseok, I'll remove it21:31
josepushed21:34
jamespagemarcoceppi, figured out my problem - old maas install suffering from nonces issue creating problems21:41
marcoceppihuh21:43
jamespagegnz21:50
jamespagelazyPower, thanks for the review21:50
lazyPowerjamespage: thanks for the high quality submission21:50
lazyPower:)21:50
jamespagelazyPower, can't take all the credit - yolanda and ivoks both worked on that one as well21:51
lazyPowerthe merge history was looonnnnggg on that one.21:51
jamespagelazyPower, fwiw that charm also works just fine on trusty as well21:53
jamespageit can deal with 12.04 rabbit and 14.04 rabbit21:53
joselazyPower: thank you :)21:54
marcoceppijamespage: good to know21:55
lazyPowerhey mbruzek22:00
mbruzekyes?22:00
lazyPoweri'm down to the new tomcat submission. Just so i'm 100% clear - this is a brandy new charm intended to deprecate tomcat6 and tomcat7 charms, correct?22:01
lazyPowernot an update to either/or22:01
mbruzeknew charm based on Robert Ayres code22:01
* lazyPower cracks knuckles22:01
lazyPowerhere we go then, let the review...COMMENCE!22:01
lazyPowerbtw ty for this <322:01
jose:P22:02
fro0gis there a Juju changelog document somewhere describing what's new in the different versions?22:07
josefro0g: I think http://changelogs.ubuntu.com/changelogs/pool/universe/j/juju-core/juju-core_1.18.1-0ubuntu1/changelog is what you're looking for22:10
fro0gjose: indeed, thanks!22:17
smarteris "juju debug-log" supposed to work with juju 1.18 and local deployments now?22:40
rick_h_smarter: I think with 1.19 now22:40
smarterok :)22:41
=== CyberJacob is now known as CyberJacob|Away
=== CyberJacob|Away is now known as CyberJacob
lazyPowerCongrats to mbruzek on his newly promulgated Tomcat charm! This is another example of a high quality submission - https://bugs.launchpad.net/charms/+bug/129571023:30
_mup_Bug #1295710: Create a new Apache Tomcat Juju Charm. <Juju Charms Collection:Fix Released> <https://launchpad.net/bugs/1295710>23:30
* jose claps23:37
davecheneylazyPower: what happens if something dies while obtain_tomcat_lock is held ?23:38
lazyPowerdavecheney: actually - good question. Its a sentinel file, so you'd have to manually release the lock.23:41
davecheneylazyPower: what does the lock, lock ?23:41
lazyPowerdavecheney: as i understand the source, it locks tomcat from doing things while its reconfiguring. so the app server continues to churn away while the hooks do their thing, aftwords it recycles the app server23:43
davecheneylazyPower: this could be fixed by running each hook in that big block in a ( ) block23:43
lazyPowerwhy "fixed"? i dont think theres anything wrong with it.23:44
=== CyberJacob is now known as CyberJacob|Away
lazyPoweri may be getting tripped up on semantics though - if that's a feature you would want out of the tomcat charm, i'd file it so the author can implement it.23:45
davecheneyif a hook fails23:46
davecheneyit'll leave tomcat locked23:46
davecheneyhow can an adminstrator see this ?23:46
lazyPowerlet me redeploy it and evaluate the behavior23:47
lazyPowerdavecheney: also thanks for the feedback23:48
davecheneylazyPower: no probs23:48
lazyPowerdavecheney: subsequent runs take control of the file descriptor lock, and if a hookf ails, its shown int eh status output23:59

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