=== al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [15:49] negronjl: mira, the moin charm got some updates a while back but we never finished it off, it looks pretty close if you want to finish off a review [16:41] jcastro: I'll look into it in a minute....id you happen to have a bug number, that will speed things up [16:41] SpamapS: I updated distcc ( https://bugs.launchpad.net/charms/+bug/944989 ). When you get a chance, can you review it again ? [16:41] <_mup_> Bug #944989: charm needed: distcc < https://launchpad.net/bugs/944989 > [16:47] negronjl: looking righ tnow [16:47] SpamapS: thx. [16:52] negronjl: *really* nice work [16:53] SpamapS: thx ... ok to promulgate now ? [16:53] negronjl: do you want to promulgate that? (don't forget to make it owned by ~charmers first) [16:53] SpamapS: Yes and yes ... [16:53] I'll +1 in the bug just for doc purposes [16:53] SpamapS: thx [16:54] SpamapS: btw ... I thought that promulgate would change the ownership of the branch to ~charmers on it's own ... [16:54] negronjl: it doesn't [16:54] SpamapS: ok .. .I'll do it [16:55] negronjl: I believe pushing to lp:charms/foo does .. [16:55] SpamapS: We'll find out very soon :) [16:55] if that doesn't do it, then I've taken to just pushing to lp:~charmers/charms/oneiric/foo/trunk with --rememeber .. then promulgating [17:09] SpamapS: distcc ... promulgated [17:09] glusterfs charm .... here I come !!!!!!!! [17:12] oooohhhhhh snap [17:13] SpamapS: hey, can you please change ownership of lp:charms/tomcat6 to ~charmers? [17:13] SpamapS: will do [17:14] negronjl: can you do the same for lp:charms/hadoop-mapreduce? that's still showing as a personal branch [17:16] SpamapS: lp:charms/tomcat6 shows as yours :P [17:16] m_3: will do [17:17] m_3: done [17:17] negronjl: cool thanks! [17:17] negronjl: doh [17:17] m_3: no worries ... [17:18] m_3: zookeeper charm appears to work OK on oneiric [17:18] jamespage: whoohoo! [17:18] m_3: should I remove the config options that are specific to precise? [17:18] its a bit misleading ATM [17:19] jamespage: I think they'll be two separate charms maintained in parallel... any better ideas? [17:19] m_3: kinda what I mean; I'll update the branch to remove the PPA options for oneiric [17:19] they can stay for the precise version [17:19] jamespage: right [17:20] m_3: or I could backport the packages in the PPA to oneiric - that would be easier actually [17:20] then people can kill themselves with 3.4.3! [17:20] jamespage: easier? really? :) [17:21] m_3: the packaging is simple so yes [17:21] jamespage: it makes sense to keep separate branches lp:~charmers/charms/oneiric/zookeeper/trunk (lp:charms/zookeeper) and lp:~cha...precise... [17:22] m_3: yep [17:22] jamespage: it's just a little backwards from the usual "direction" of the branch :) [17:22] m_3: charm backport - I like it! [17:22] * m_3 groan [17:23] m_3: OK if I promulgate that now! [17:23] ? [17:23] jamespage: just still scared of backports in packages :) [17:23] jamespage: please go ahead and shrink wrap it [17:24] m_3: https://code.launchpad.net/~charmers/charms/oneiric/zookeeper/trunk [17:24] deone [17:24] done even [17:25] jamespage: awesome! [17:25] charms are almost tended... jcastro will even be happy... we're net positive count after the changes :) [17:26] \o/ [18:40] What I do when a bootstrap instance state is not-started? [18:41] check the log on the server in /var/log/juju/ [18:43] avoine: -bash: cd: /var/log/juju: No such file or directory [18:43] =/ [18:43] if there is none and you are on aws also check /var/log/cloudinit/ [18:44] andrewsmedina: what provider are you using ? [18:44] SpamapS: ec2 (openstack) [18:45] using default juju version works fine, but with the last ppa version, bootstrap returns "not-started" status [18:45] bootstrap usually doesn't return anything.. can you explain? [18:46] SpamapS: after bootstrap, when I execute juju status, returns the bootstrap instance status [18:49] andrewsmedina: check in /var/log/cloudinit/ if something have failed there [18:50] andrewsmedina: right, ok. So you need to use the EC2 API tools to figure out why the instance hasn't started yet. [18:50] andrewsmedina: euca-get-console-output on the instance ID should help [18:51] its probably the same problem of the metadata server not being configured to return values for 1.0 api version [18:51] its a configuration (.ini) fix for nova [18:52] ie. bug 942868 [18:52] <_mup_> Bug #942868: metadata service reports to support 1.0 but serves incorrect data < https://launchpad.net/bugs/942868 > [18:53] Right! [18:53] it feels like a packaging bug, we could update juju to just use latest or another common version that's enabled by default [18:53] major DOH there ;) [18:53] ie. work around it [18:53] SpamapS: is everything ok in the instance [18:53] * SpamapS is getting giddy.. in 2 hours his upstream will go from 1Mbit to 5Mbit .. woot. [18:54] hazmat: I think we should update juju to use the *latest* common version that we know we can support. [18:54] SpamapS, technically the metadata we're after is common to all of the extant versions [18:55] SpamapS: but I think that juju does not found zookeeper agent in this instance [18:55] hazmat: isn't there a "best effort" url? [18:55] it looks like 2009-04-04 is the latest version available [18:55] SpamapS: cloud-init boot finished at Mon, 05 Mar 2012 18:53:55 +0000. Up 300.53 seconds [18:55] SpamapS, 'latest' is latest, but some risk of deviation.. [18:55] pretty minor though [18:55] internally these all dispatch to the same implementation for ostack [18:56] andrewsmedina: right, so from there, you need to look at the juju agent logs,which will be in /var/log/juju .. or perhaps if they had catastrophic errors, in /var/log/upstart [18:56] the error should be in the console output [18:56] hazmat: try:;latest;except:;try:known_old_version [18:56] hazmat: ok so now /charms says 60 official [18:57] jcastro, woot! [18:57] m_3, nice [18:57] hazmat: heh ok, so we're sure this is right right? [18:57] SpamapS: but with an old juju version works fine [18:57] I still see some are counted twice [18:58] jcastro, /charms is cross series, add a series to it if you want a particular series [18:58] jcastro, ala http://charms.kapilt.com/charms/oneiric [18:58] oh cool! [18:58] got it! [18:58] ok, so 54! [18:58] hazmat: with the agents being started in upstart jobs.. are you sure the error would be in the console? [18:59] SpamapS, yes, the metadata server error happens pre-agent during juju-admin initialize [18:59] which is just a cli command for cloud-init [19:00] andrewsmedina, can you pastebin the console output? [19:00] hazmat: well at least we have a fail early then. ;) [19:06] SpamapS: http://dpaste.org/u7oDI/ [19:08] hazmat: ^^ [19:09] andrewsmedina, that's status output.. i meant the equivalent of ec2-console-output.. or the cloud init log [19:11] but that's the output i would expect for this bug from status [19:11] hazmat: http://dpaste.org/1qhg4/ [19:11] huh.. [19:12] everything seems ok [19:12] andrewsmedina, if your logged on the machine.. what's the output of.. curl http://169.254.169.254/1.0/meta-data/instance-id [19:14] hazmat: humm [19:14] hazmat: http://dpaste.org/oRDnk/ [19:17] hazmat: juju-sample groups isn't created [19:19] hazmat: im wrong [19:19] juju-sample groups is here [19:25] hazmat: the problem is my http_proxy =/ === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away [19:39] Ugh always the proxy [19:47] errand time, bbiab [20:00] I removed the proxy but nothing has changed =/ [20:21] negronjl: nice work on the review! [20:21] so we can promulgate that badboy? [20:21] jcastro: sure [20:22] jcastro: checking to see if he has access to do it. If he doesn't, I'll merge it. [20:22] I don't think he does [20:23] he looks like a new contributor [20:23] time for a tshirt and a mug! [20:26] jcastro: python-moinmoin merged and pushed and promulgated [20:26] * jcastro rings the bell [20:26] clear [20:26] third one in 2 days fellas! [20:26] jcastro: do we need to upgrade this guy ? [20:27] what do you mean? [20:27] jcastro: how many charms does patrick have ? [20:27] this is his first [20:27] jcastro: nm then ... ignore me [20:28] your mixing him up with Nathan I think [20:28] which has like 4 or 5 now or something [20:28] but only interested in writing, not reviewing [20:44] SpamapS: got time for a quick G+? [20:44] I need like 45 seconds for you to explain something to me [20:53] m_3: https://launchpad.net/~chris-lea/+archive/node.js [20:53] FYI I didn't know there was a node PPA [20:59] <_mup_> Bug #947483 was filed: Support for container relations < https://launchpad.net/bugs/947483 > [21:00] bcsaller, lbox needs the bug flag to show up in the kanban [21:01] hazmat: and I think rerunning it isn't adding the bug for me now, but I think it used to [21:02] bcsaller, i just did it by hand.. [21:08] <_mup_> juju/force-upgrade r458 committed by kapil.thangavelu@canonical.com [21:08] <_mup_> some environment destruction tests where missing inlinecallbacks decorator [21:19] niemeyer, the charm test spec looks like it could use another review.. https://codereview.appspot.com/5624044/ [21:21] m_3, what's the url for your charm testing jenkins instance? [21:22] * hazmat keeps forgetting [21:33] jcastro: back from lunch, we can G+ any time. [21:42] hazmat: it's moving ec2 accts atm... gimme a few [21:54] hazmat, can you respond to julian edwards and gavin panella re the maas provider branch? [21:56] * SpamapS still wishes providers had been made fully pluggable so they would be useful without being fully integrated into juju.environment.config :-P [21:57] jimbaker, sure [21:57] hazmat, thanks! [22:00] hazmat: my problem is that juju version in bootstrap in different from my machine [22:01] hazmat: in my machine I use ppa version [22:01] andrewsmedina, juju should detect that and use the ppa version when bootstrapping, unless a different juju-origin has been set in environments.yaml [22:03] SpamapS, good point. a role for ju-jujitsu? or is that just crazy talk? [22:04] hazmat: I defined juju-origin: ppa, but the version installed is the distro version [22:04] jimbaker: totally crazy. juju has to adapt to make them pluggable [22:05] SpamapS, after the cli plugins got nixed i gave up on out of tree plugins [22:05] SpamapS, its moot at this point... and in the future juju will be a binary cmd [22:05] big fat juicy binary [22:06] hazmat, :) [22:06] can still do plugins [22:06] SpamapS, for providers.. maybe.. but they'd be out of process.. i wouldn't hold your breath [22:11] hazmat: go has no dlopen? [22:14] SpamapS, yeah.. it has LoadLibrary.. [23:24] hey, I'd like to test out my charm using lxc containers I saw mentioned a while back. Does this work, and is there some documentation on how to do it? [23:36] SpamapS: where's the latest charm testing spec branch? [23:36] that moved into juju itself right? or is it still in charm-tools?