=== schwuk_away is now known as schwuk [09:35] if I want to add 20 units to postgres do I type "juju add-unit postgres" [09:36] gary_poster: rick_h Makyo ^ [09:36] juju add-unit -n 20 postgres I think [09:36] luca___, juju add-unit postgres -n 20 [09:37] luca___, Or either works. The docs are conflicting :P [09:37] Makyo: I see, thank you [09:40] someone needs to juju add-unit luca's brain -n 20 [09:48] If I could scale up like that, I imagine this would all be a lot easier. [10:06] hatch, http://www.youtube.com/watch?v=mC4Z8uIoVcQ [10:06] For whenever. [10:21] rick_h: http://10.20.69.166:2464/ === schwuk is now known as schwuk_away === schwuk_away is now known as schwuk [11:02] huwshimi: Scale up with these constraints? [13:06] Does anyone remember writing this comment in the inspector CSS? "The height of each list is 35px so setting this to help the auto max height script." [13:11] huwshimi: I think that is from the auto adjusting height stuff right? [13:14] rick_h: If you get a chance could you come over and explain what's going on there? [13:15] huwshimi: yea [13:15] rick_h: No hurry [13:15] yea, trying to figure out if we're having an 'all' meeting or not [13:21] frankban, categoryWrapperNode.filter(function(d) { return d3.select(this).classed('landscape-needs-reboot'); }) [13:32] jcsackett, hangout? [13:41] bac: I finished my Makefile branch: https://code.launchpad.net/~benji/charmworld/makefile-tweaks/+merge/184105 [13:41] I would enjoy a review if you have time. [13:42] benji: ok. i'm reviewing abel's branch atm [13:42] thanks [13:42] benji: my icon2 branch is ready too. i'm going to kill the initial MP [13:42] bac: I'll review it [13:42] (just let me know when and where) [13:55] If anyone wants to review my charmworld redesign branch it's here: https://code.launchpad.net/~huwshimi/charmworld/charmworld-styling/+merge/184078 [13:56] Don't worry, it's only a 11,237 line diff. [14:00] huwshimi: I'll take a look. [14:00] If I'm not back in three weeks tell my family that I love them. [14:01] frankban, https://gist.github.com/makyo/6450487 [14:02] benji: Haha, thanks. My apologies. [14:02] I wonder if I can get this diff up on Rietveld. It would be nice to do this as an in-line review. [14:13] benji: makefile branch approved but with one question about use in production. [14:13] cool, I'll look [14:14] gary_poster, review done [14:15] thank you I saw hatch :-) [14:15] benji: looks like huwshimi's branch breaks the makefile wrt css [14:15] bac: yeah, I was just noticing the same thing [14:15] benji: quick, land yours and maybe that'll fix it [14:16] heh, I doubt it [14:16] bac, benji: Oh did I not remove some bootstrap paths or something? [14:17] hatch: can you QA my branch that gary reviewed please? https://codereview.appspot.com/13368049/ [14:18] huwshimi: there is a rule that depends on theme.less but it has been removed [14:19] oops [14:19] huwshimi: so is the css all static now? not less and not generated? [14:20] bac: Yeah, in fact we just load the CSS hosted on ubuntu.com [14:20] eww, the charm is written in shell script [14:21] hatch, https://codereview.appspot.com/13563043 [14:23] sinzui: so, juju is now throwing "found no instances" i assume because i'm no longer shuttled. however, nova lists nothing for me as well. do we have nothing live, or is my nova lying? [14:25] jcsackett, you are being lied to. There nova list shows me 6 staging machines, 2 qa machines and a bac instance [14:25] sinzui: ok, so juju is probably configured right, but not canonistack itself. interesting. [14:27] This morning I source orange stack, set JUJU_HOME to the dir to my env.yaml, ran sshuttle, did juju status [14:29] huwshimi: If you can fix that make issue it would help me in the review (I'm trying to get the diff up on Rietveld so I can do in-line comments) [14:30] benji: Sure, will take a look now. [14:30] cool [14:30] sinzui: right, i was going to shuttle, but i don't want to ping you every time i need to get the IPs to do so. so time to debug nova. [14:43] huwshimi: thanks for the rework of the charm and bundle pages. sorry they were such a mess to start. [14:46] adeuring: may i assign the ES test problem card to you on the kanban board? [14:47] bac: well... not sure if this fits into the tasks I am supposed to works on in the next time. sinzui: ^^^ [14:47] adeuring: i figured your current branch did the work. [14:48] adeuring: i'm just trying to clean up the board and thought you should get credit for work already done. [14:48] bac: no, not at all. I suspect that we need to not only use the ES call health() but to also check the index status, as descriibed here http://www.elasticsearch.org/guide/reference/api/admin-indices-status/ [14:48] orangesquad: Here's a blog post that I believe describes how the launchpad jenkins plugin is used. http://qualityhour.wordpress.com/2012/06/14/continuous-integration-with-jenkins-and-launchpad/ [14:48] benji: i think you can just remove the CSS rule from huwshimi's makefile. the css is all static now. [14:49] bac: he's looking at it; I figured he would know what was up better than I [14:50] bac: the challenge with this APU call is to figure out what information is relevant. And perhaps also, _if_ this URL provides relevant information... [14:51] bac: No problems at all! [14:52] benji: I've got caught up here a bit. I think you can probably remove any/all css rules there, but I was going to grab Rick to make sure I wasn't breaking anything, but he's caught up with something now. [14:53] huwshimi: ok, I'll take a look [14:54] bac: ah, i misunderstood your question. I thought you refered to email. Yes, the pending MP should fix some problems that occur even when the ES server is "stable". [14:59] hey gary_poster [14:59] I found an interesting thing [15:00] https://chinstrap.canonical.com/~liam/cassandra-setup.jpeg [15:00] I asked the guys in IS to make me a diagram on how we do cassandra for a blog post [15:00] and this is what they came up with [15:00] but like the GUI doesn't show units and relationships together, so the diagram looks rather lame in comparison [15:00] I know we don't really target "make it look impressive when there's a ton of units" as a use case [15:01] but I thought I'd point it out [15:02] hatch, https://codereview.appspot.com/13515044 [15:10] jujugui Review please for flagging a service as charm_changed https://codereview.appspot.com/13560043/ [15:17] jcastro, hey. Fair enough. :-) Yeah, I don't expect we'll get a unit view any time soon. [15:29] sinzui: now that staging is up and happy, i'd like to get production on the same rev 381. anything i should know besides just following your google doc? [15:29] bac: Nothing special today [15:30] okey doke, off to file an rt [15:31] bac: I ponder making staging again at EOD because I think we want to use a floating IP. I will also need to land a branch to make migration 18 safe for fresh deploys. [15:42] benji: Did you have any luck with the makefile? [15:42] huwshimi: yep, tests running now [15:43] benji: Great! [15:43] huwshimi: tests passed [15:43] I'm going to try to get this up on Rietveld now. [15:43] benji: Have fun! [15:43] benji: I think 10,000 lines of that review are removing bootstrap, so it's not as bad as it seems. [15:44] benji: are you going to be able to get to my bundle-icon branch or should i ask an orangish person? [15:44] bac: oh, I didn't know it was ready. I asked for you to tell me where it was when it was ready. [15:44] * benji looks at the backscroll to see if he missed something. [15:44] benji: yeah, it's been ready. sorry i didn't give you an url...its in the normal spot [15:45] benji: https://code.launchpad.net/%7Ejuju-jitsu/charmworld/trunk/+activereviews [15:45] thanks [15:46] benji: only 244 lines of that are code, the rest svg [15:46] cool, looking now [15:56] jujugui https://codereview.appspot.com/13566044 quick review please, I broke zooming :( [16:07] bac: the branch looks good [16:07] thanks [16:14] bac: I added a card for the docstringageddon [16:15] benji: great. we're much better at making cards than finishing them. :( [16:15] * bac points finger at self [16:15] heh [16:16] benji: no call today, right? [16:16] as far as I know [16:16] nice. /me -> lunch === schwuk is now known as schwuk_away === schwuk_away is now known as schwuk [17:06] jujugui quick fix for services appearing to jump on drag https://codereview.appspot.com/13547045 [17:09] Makyo: I did a review (no QA) === schwuk is now known as schwuk_away === schwuk_away is now known as schwuk === schwuk is now known as schwuk_away === schwuk_away is now known as schwuk === schwuk is now known as schwuk_away [17:54] hi sinzui, fyi production has been updated https://pastebin.canonical.com/96998/ [17:55] we should have a charm icon contest. i vote bip for most engaging. http://manage.jujucharms.com/charms/precise/bip === schwuk_away is now known as schwuk [19:03] hi benji, there is a problem with your makefile for the run target. es_update instead of es-update. i've fixed it in my branch. === schwuk is now known as schwuk_away [19:10] bac: ooh, thanks; I'll sneak a quick change into trunk [20:26] benji: something is clearing out my local db between runs. could it be a change to the makefile? [20:26] benji: could be an undesired behavior in production. [20:26] bac: I don't see how. [20:27] bac: between runs i.e., "make run"? [20:27] benji: perhaps 'make test' / 'make run' cycles. [20:31] benji: a mostly mechanical change: https://code.launchpad.net/~bac/charmworld/singularize-bundle/+merge/184181 [20:32] bac: I'll take a look. Re. db clearage: I just did a "make run" and checked that the apache charm page worked, did a "make test" (the tests passed), and then another "make run" and the apache page was still working [21:00] thanks for the review benji [21:01] my pleasure [21:07] benji: it is ES that is getting blown away. do a run / test / run and then do a search for 'precise' [21:07] benji: i wonder if it is es-update. it sure takes a long time now [21:07] * bac dogs walk [21:09] bac: yep, I think I have the same thing happening, all searches return no results