[00:03] <manjiri> Hello. I am running into bug https://bugs.launchpad.net/charms/+source/postgresql/+bug/1239681 which appears to have been fixed by sinzui. How can I get that fix?
[00:03] <_mup_> Bug #1239681: relation-get failing with 'permission denied' <regression> <juju-core:Triaged> <postgresql (Juju Charms Collection):Triaged> <https://launchpad.net/bugs/1239681>
[00:05] <marcoceppi_> manjiri: what version of juju are you using?
[00:07] <manjiri> marcoceppi_: ii  juju-core                                    1.16.6-0ubuntu1~ubuntu12.04.1~juju1              Juju is devops distilled - client
[00:10] <marcoceppi_> manjiri: also, it doesn't say it's been fixed, merely triaged for 1.18
[00:11] <manjiri> I think I recently upgraded juju-core. Do you recommend that I down-grade? To which?
[00:14] <hazmat> manjiri, can you reliably reproduce that?
[00:15] <hazmat> manjiri, are you running into that  with the postgresql charm?
[00:16] <manjiri> hazmat: I am not running postresql charm. I am developing my own (set of) charms for deploying Contrail software. I think the symptoms are the same.
[00:17] <hazmat> manjiri, namely that relation-get tosses an error?
[00:17] <manjiri> hazmat: yes. that is correct.
[00:18] <manjiri> hazmat: as to whether I can reproduce the problem - I am seeing it now. I was thinking of rebooting the machine in the hope that the problem would go away.
[00:18] <hazmat> manjiri, could you send the log for the unit in /var/log/juju/
[00:18] <hazmat> i'll try reproducing w/ the postgres charm | unit test
[00:19] <hazmat> i'm on 1.17/trunk though
[00:19] <manjiri> hazmat: I am new to IRC. When you say "send" the log - what do you mean?
[00:20] <hazmat> manjiri, typically pastebin.ubuntu.com if its not sensitive
[00:20] <sarnold> (the pastebinit tool is awesome for this)
[00:20] <hazmat> manjiri, welcome to irc btw.. also there's a command line client
[00:20] <hazmat> sudo apt-get install pastebinit
[00:21] <hazmat> you can point it to a file as a param or pipe data into the pastebinit cli
[00:22] <marcoceppi> hazmat: --wait in deployer just subscribes to the event stream from the API, right? It doesn't actually do proper "wait for 'idle' environment", does it?
[00:22] <hazmat> marcoceppi, the latter doesn't exist
[00:22] <hazmat> atm
[00:22] <hazmat> marcoceppi, yes.. its event stream wait and configurable timeout
[00:22] <marcoceppi> hazmat: didn't think so, that's why I got kind of excited when I saw the option
[00:22] <hazmat> marcoceppi, and stop/shortcircuit on error
[00:23] <hazmat> rephrasing its wait for x seconds without error post action
[00:23] <manjiri> hazmat: Please check out http://pastebin.ubuntu.com/6991890/
[00:24] <hazmat> manjiri, are you changing users in your charm?
[00:25] <hazmat> manjiri, the juju cli commands use a  unix socket owned by root. changing user would be the most likely cause.
[00:26] <manjiri> hazmat: no change in user
[00:27] <manjiri> hazmat: to which cli commands are you referring?
[00:27] <hazmat> manjiri, ok
[00:27] <hazmat> manjiri, all the commands used to interact with juju from a hook (relation-ids, relation-list, relation-get, relation-set, config-get, open-port, close-port, etc)
[00:28] <manjiri> hazmat: got it. to reconfirm, my charm does everything as 'root'
[00:34] <manjiri> hazmat: do you have a recommendation? Downgrade juju-core?
[00:37] <hazmat> manjiri, not atm
[00:37] <hazmat> still trying to understand root cause, but past eod, so intermixing other things
[00:39] <hazmat> manjiri, 1.14 is very old.. not recommened
[00:40] <manjiri> hazmat: I have 1.16.6-0ubuntu1~ubuntu12.04.1~juju1
[00:40] <hazmat> understood
[00:40] <hazmat> manjiri, is your charm opensource?
[00:40] <hazmat> if so where can i find it..
[00:40] <manjiri> hazmat: not yet
[00:40] <hazmat> currently trying to run the postgresql charm unit tests per the bug report
[00:41] <manjiri> hazmat: I can provide code snippet if you think that it will help
[00:44] <hazmat> manjiri, that would be helpful
[00:46] <manjiri> hazmat: Is it OK if I email it to you?
[00:46] <hazmat> manjiri, sure
[00:47] <manjiri> hazmat: Give me a few minutes
[00:47] <hazmat> manjiri, np
[00:48] <hazmat> stub, ping.. how do you run these postgresql charm tests
[01:29] <manjiri> hazmat: Sorry that took so long, but I have sent you the information in an email. I will reboot the machine to see if I can continue to make progress without being hit by this bug
[01:33] <hazmat> manjiri, got it.. i'll have to look at in the morning its late here.. and i'm done for the night
[04:54] <stub> hazmat: Bootstrap an environment, then 'make test'. tests/00_setup lists packages that might be needed
[04:54] <stub> hazmat: it has been a while since I tried anything but the local provider
[10:08] <hazmat> stub, thanks
[10:09] <hazmat> stub, i use a variation a manual provider variation on local provider with lxc btrfs snapshots.. should be about the same.
[10:11] <hazmat> stub, you still have that failing test per the bug report .. http://pad.lv/1239681
[10:11] <hazmat> ?
[18:41] <danob> i installed LXC container in my ubuntu desktop and when i start my desktop or restart my desktop, why i need to run a2ensite default everytime inside apache2/0 unit???
[18:51] <marcoceppi> danob: you shouldn't are you setting a vhost template?
[19:02] <chris38home> hi, I'm trying local provider with fresh checkout and juju add-machine lxc:0 stays in pending mod, is there any prerequisite to install ?
[19:09] <marcoceppi> chris38home: uh, I don't think you can LXC on LXC
[19:12] <sarnold> you can do nested lxc: https://www.stgraber.org/2013/12/21/lxc-1-0-advanced-container-usage/
[19:12] <chris38home> ok then juju add-machine should create a new lxc container ?
[19:12] <sarnold> no idea how well juju will handle it though :)
[19:13] <chris38home> non nested is also pending
[19:18] <marcoceppi> chris38home: well, non-nested needs to be instarted before nested can work
[19:18] <marcoceppi> chris38home: can you show your juju status in a pastebin?
[19:28] <chris38home> marcoceppi,  http://pastebin.com/AB8HDfZS
[19:43] <chris38home> maybe I have a problem with routing : cannot load index "http://10.0.3.1:8040/tools/streams/v1/index.sjson"
[20:21] <jcastro_> is this bucket thing still a thing? http://askubuntu.com/questions/425745/maas-juju-bootstrap-404
[20:21] <jcastro_> I haven't seen that one in a while
[20:52] <Lord_Set> Hello everyone
[20:52] <marcoceppi> o/ Lord_Set
[20:55] <Lord_Set> How is the world of Juju going?
[21:05] <marcoceppi> Lord_Set: really well, imo
[21:06] <Lord_Set> Glad to hear it. I am really enjoying using it as an app deployment platform. I am doing a pretty good sized test today with it.
[21:08] <Lord_Set> Going to be using MAAS and Juju to deploy a rack of 35 servers with Openstack and then a second rack of 20 servers to build a Hadoop Cluster. Also going to doing some multi-cluster controller between different sites tests. The locations are connected via VPN tunnels.
[21:08] <Lord_Set> We are upgrading to 100mbps burstable to 1gbps fiber soon between locations.
[21:08] <marcoceppi> Lord_Set: Sweet! Let us know if we can help in any way
[21:10] <Lord_Set> I will for sure. I got an email from Dustin Kirkland this morning saying he wanted to discuss what my team is doing with MAAS, Juju and Ubuntu. Was after a good and long conversation with Julian Edwards of the MAAS tema.
[21:10] <Lord_Set> err team
[21:12] <jcastro_> Lord_Set, this is awesome, let us know how it goes for you
[21:13] <jcastro_> marcoceppi, I hate to sound like a raving nut, but putting the docs on github makes things so easy now with the inline editing.
[21:13] <marcoceppi> jcastro_: you sound like a raving nut ;)
[21:14] <Lord_Set> I will
[21:14] <jcastro_> Apache Syncope charm incoming to the queue!
[21:15] <Lord_Set> I am hoping with this tech startup I am working with to form an awesome partnership with the MAAS and Juju teams. We have a lot of resources and big plans for deployment with 2 different tech startup companies we are moving forward with.
[21:16] <Lord_Set> Currently 4 geographical sites with a total of 12 racks currently. 2 of which are at Switch SuperNAP in Vegas
[21:16] <jcastro_> negronjl, I kind of need bug 1267222 fixed, could you possibly update it with what needs to be done and I can ask someone else to fix it?
[21:16] <_mup_> Bug #1267222: Race condition when deploying a simple replica set <audit> <mongodb (Juju Charms Collection):New for negronjl> <https://launchpad.net/bugs/1267222>
[21:17] <negronjl> jcastro_, sorry about that ... I'll update today
[21:17] <marcoceppi> jcastro_: didn't lazyPower fix that ?
[21:18] <jcastro_> I dunno, did he?
[21:18]  * marcoceppi waits for lazyPower
[21:19] <lazyPower> Nope, i fixed a race condition in mediawiki, and pointed it out back in the day
[21:19] <lazyPower> but have not applied any patches to mongodb charm
[21:37] <jcastro_> lazyPower, marcoceppi, mbruzek: hey when you guys get time can you go through and vote in the juju tag?
[21:37] <lazyPower> negronjl: I do however have tests in amulet to assert that it works - i think you already patched it.
[21:37] <jcastro_> dimiter is answering a ton but not really getting any upvotes
[21:38] <lazyPower> negronjl: latest bzr branch is lp:~lazypower/charms/precise/mongodb/ci-fix
[23:16] <jcastro_> mbruzek, marcoceppi: heya, looks like wildfly is ready to go
[23:16] <jcastro_> what do you say about some promulgation?
[23:17] <marcoceppi> jcastro_: needs a finaly review, but otherwise, sure
[23:17] <mbruzek> Yeah it needs to be reviewed by a charmer.
[23:17] <mbruzek> Marco let me know if you have any questions or find anything that I did not.