=== thumper-gym is now known as thumper === timrc-afk is now known as timrc [02:40] guys, is it fine if a non-charmer reviews a charm pointing some errors? I'd like to do some review when I have some time [02:41] jose: yes, absolutely [02:41] many hands make light work [02:41] cool, thanks! :) === vladk|offline is now known as vladk === vladk is now known as vladk|offline === vladk|offline is now known as vladk === vladk is now known as vladk|offline === vladk|offline is now known as vladk === CyberJacob|Away is now known as CyberJacob === rogpeppe is now known as rogpeppe1 === rogpeppe1 is now known as rogpeppe [12:53] Good morning, all. [12:59] morning cory_fu [13:37] We have a system that juju status tells us that 2 system's agent state is "down" [13:37] Is there a way to restart the juju agents on a system? [13:54] mbruzek: upstart should handle restarting the juju-agent [13:55] I've brought several machines up and down this weekend that are manually provisioned. There is an upstart task for resuming operation -- i dont know if they fail out if the bootstrap node is unavailable... i didn't look that closely at it. [13:55] lazyPower, juju status shows 2 agents down, so presumably the upstart is not working. Can we manually restart them? [13:56] should be able to just call the service back up. let me remote into a jujub ox and peek at /etc/init [13:56] mbruzek: jujud-machine-1.conf exists [13:56] that should be the agent. so service jujud-machine-# restart should fix it [14:02] lazyPower: ever seen this error: [14:02] ERROR state/api: websocket.Dial wss://10.0.3.1:17070/: x509: certificate signed by unknown authority... [14:02] http://pastebin.ubuntu.com/7216718/ [14:02] Yeah, it was bootstrapped by another workstation and your client is bugging out over the ssl signature [14:03] this is why we actively suggest jumpstations for working on remote systems as a team presently [14:03] i don't see how that's possible, this is local [14:04] did you wipe your juju config between bootstrap and now? [14:04] those are the only 2 cases where i've seen that. [14:04] you mean environments.yaml? [14:04] nah, i mean the jenv stuff [14:04] i don't even know where that is [14:05] so i don't think i deleted it [14:05] any way to work around this? [14:05] tvansteenburgh: it's in ~/.juju/environments [14:05] tvansteenburgh: you have to tear down and start again [14:05] ^ [14:05] if i force destroy the env then rebootstrap, i get this error running juju status [14:05] tvansteenburgh: that's, interesting [14:06] i seem to do all sorts of interesting things [14:06] tvansteenburgh: destroy again [14:07] rm -f ~/.juju/environments/local.jenv [14:07] try again [14:08] after destroying, there's nothing in ~/.juju/environments [14:10] the full error is: [14:10] ERROR 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\"") [14:10] filed a bug on it this morning but i'm really just looking for a way to get passed it now [14:11] tvansteenburgh: destroy, again, run bootstrap with --debug flag [14:11] paste output [14:13] marcoceppi: http://pastebin.ubuntu.com/7217166/ [14:13] uhhh [14:13] and now juju status works fine [14:13] you're welcome \o/ [14:13] ;) [14:14] \o/ [14:15] bonus [14:16] hi marcoceppi, do you have any plan to update the wordpress charm? [14:17] overm1nd: I've had plans to update that charm for over a year [14:17] overm1nd: it's the time I don't have, did you want something done? [14:17] ahaha I can feel the pain [14:17] I was interested in multisite support [14:17] overm1nd: yeah, I have a plan for that, which would work nicely with juju [14:18] overm1nd: I may try to attack it this weekend, unless you wanted to give it a go [14:18] I tried but it feels too complex for me [14:19] I also tried to start a new charm but I can get the sourse form charm get command [14:19] from* [14:19] I know I don't need it but something is vroken in my env I need to look at it [14:19] to tell the true I'm also evaluating puppets and ansible [14:23] any of you ever ran into these when mining: http://paste.ubuntu.com/7217201/ ? I just started a priate network server and all my miners get that, trying to think of why? [14:24] oops wrong chat [14:30] hazmat, responding to a friday ping now. you still need anything? [14:30] smoser, mr wolf was calling in reinforcements at the time, i think its all clear atm. [14:31] well, yeah. [14:32] smoser, how was the game? [14:39] game was cold (very strong gusts of wind, ~40 degrees F and ~30F with windchill). and cubs lost. but i was at wrigley field for opening day, so that part was awesome. [14:54] hello everyone, I'm having an issue with juju deploy that is showing up since I updated juju-core to 1.18. Error is: "only charm store charm references are supported, with cs: schema" when deploying the solr-jetty charm [14:54] more debug output in https://pastebin.canonical.com/107885/ [14:54] any ideas? [14:55] perrito666, would you know who can help me with that ^? === Darkwing_ is now known as Darkwing [15:32] mbruzek: Hi, if you've time, can you check this charm-helpers MP: https://code.launchpad.net/~michael.nelson/charm-helpers/fresh-ansible-relations/+merge/214203 [15:32] mbruzek: bloodearnest has already checked it and is +1 (hrm, bloodearnest, can you add your vote there when you get a chance) [15:32] Yes I will have a look [15:32] mbruzek: he's the other ansible-support user besides me :-) [15:32] Thanks. [15:47] noodles775: done [15:55] having trouble setting up juju to run disconnected from the internet on openstack. I did a juju sync-tools --all --local-dir=/tmp/juju on a connected machine, moved the result to a web server accessible by the openstack system [15:56] then I created an "images" diretory on the web server and grabbed releases/precise/*12.04*amd64* (the only vesion I care about for this deployment) [15:56] also grabbed cloud-images release/streams/v1/index.json and com.ubuntu.cloud:released:download.json [15:56] sinzui, _THANK YOU_ for such finely detailed release notes [15:56] that is an epic announcement [15:57] removed all the entries which weren't 12.04 amd64 [15:57] added "tools-metadata-url: http://server/juju/tools" and "image-metadata-url: http://server/juju/images" to my environments.yaml [15:58] juju -v bootstrap gives me "ERROR bootstrap failed: cannot start bootstrap instance: invalid URL "http://cloud-images.ubuntu.com/release/streams/v1/index.json" not found" [15:58] what am I missing? [15:59] tclarke, /release/ is missing a "s", /releases/ [16:00] thought of that...I have a symlink between "release" and "releases" [16:00] tclarke, you may not need to set that since that is the default location to search [16:00] so both exist [16:01] here's the full directory listing served by the web server: http://www.pastebin.ca/index.php [16:01] hang on, wrong link [16:02] here, pastbin.ca doesn't want to work for me :) http://pastebin.com/eVP12L8q [16:03] MAAS / nodes have limited internet access after install | http://askubuntu.com/q/444563 [16:05] tclarke, oh, yes, restricted networks must open access to that machine or mirror the images locally. [16:06] sinzui: I release that...that's what I'm trying to do but juju is ignoring my image mirror config and is still trying cloud-images [16:07] tclarke, right, and image-metadata-url is the right option to change :( [16:07] hmm - it's only the MaaS provider that can place use --to lxc:n ? [16:07] is there an equiv of "juju sync-tools" for the images? [16:08] tclarke, I configure tools-metadata-url most of the time and it honoured. There is no effort to talk to streams.canonical.com when I change tools-metadata-url [16:08] or a similar tool...I suspect I'm either missing something in the image repo or it's not properly setup but there seems to be no documentation on creating a mirror [16:08] tclarke, maybe there is a jenv file present...juju is ignoring the yaml file. [16:09] tclarke, you may need to delete $JUJU_HOME/environments/my-env.jenv [16:09] I've attempted "rm -rf .juju/environments" a couple times...didn't seem to help [16:09] tclarke, juju does not warn you that it found a cached config and is using it [16:11] jamespage, do you have any experience with image-metadata-url being ignored with maas configs as tclarke reports? [16:11] not maas tho, openstack [16:30] progress....I seemed to have been missing a level in the directory tree.. [16:30] when I fix that, I get a different error about an invalid index.json. [16:30] my index.json is: http://pastebin.com/vRc6USiS [16:58] My MAAS import boot images never finishes.. any idea how I can check the status of it? [16:58] something is broked [16:58] looking at /var/log/maas [16:59] nothing good there [17:00] well ok I get this [17:01] who is the resident mysql charm expert? [17:01] i have a master:slave setup, and the reads are going to master and writes to slave [17:01] http://paste.ubuntu.com/7217896/ [17:02] does that error make sense to anyone [17:02] Invalid: The input field 'username' was not expected. === mramm2 is now known as mramm [17:38] argh [17:52] tvansteenburgh: that would be marcoceppi === vladk is now known as vladk|offline === BradCrittenden is now known as bac [19:24] jose, ping [19:31] jcastro: pong [19:31] hey if a charm doesn't have a logo [19:31] you don't need to make an icon for it [19:31] example: pictor [19:31] oh, pictor, dustin made those [19:32] I emailed him and asked and he replied with the logos [19:32] oh well you can remove them [19:33] huh? [19:33] why's that? === lazyPower changed the topic of #juju to: Weekly Reviewer: bcsaller || Welcome to Juju! || Docs: http://juju.ubuntu.com/docs || FAQ: http://goo.gl/MsNu4I || Review Queue: http://goo.gl/9yBZuv || Unanswered Questions: http://goo.gl/dNj8CP === Ursinha_ is now known as Ursinha [20:16] jose, it's better to just show a blank icon than a made up one with no branding [20:16] oh, you meant for the ones that don't have an icon [20:16] it's literally an orange box with text, it looks like it was made by a developer [20:16] :P [20:16] I'll make sure to fix that asap [20:16] you know what, I'll do it now [20:17] but pictor should be good to go, I think [20:17] yeah, either a totally awesome icon that matches upstream's brand, or a template [20:17] marco had some other feedback [20:17] I'll wait for it, then [20:51] jcastro: marcoceppi: I have a django charm that uses gunicorn, but I need something to service static files, what's the right path for this? [20:52] I should also note that it currently uses multiple gunicorn nodes behind an haproxy [20:52] and currently the static files are references with relative URLs like /static/css/site.css === CyberJacob is now known as CyberJacob|Away === CyberJacob|Away is now known as CyberJacob [21:29] mhall119: probably co-locate apache2 and gunicorn on the same machine, or nginx [21:30] and use the web daemon as a reverse proxy to the django app, and used for serving the static assets [21:31] i dont have much experience with gunicorn charm however - so this may be a bad suggestion... but thats my knee jerk reaction to the question. [21:33] jcastro: can I use swift-dependent things when deploying to my local LXC for testing? === ubot5` is now known as ubot5 [21:40] mhall119: we have 2 approaches to servcing static resources: [21:40] 1) a custom subordinate charm that builds/tars/deploys the assets on a service's units, then set the service's vhost config (apache in our case) to serve the static assets directly [21:41] 2) a deploy to a swift bucket and then a proxy through apache to swift [21:41] 2 is nicer, and less work [22:21] bloodearnest: do you have an example config for 2? [23:00] mhall119: apache2 === CyberJacob is now known as CyberJacob|Away [23:27] mhall119: I do't, we're just trying that approach now - but noodles775 can probably help [23:27] *don't === arosales_ is now known as arosales