[10:29] A pox on double bookings. [10:41] hi gmb [10:41] Hi frankban [10:42] gmb: can you remember me why you replaced addSkip with addError in zope.testing formatter.tear_down_not_supported? [10:43] frankban, Because we thought it was The Right Thing. But it didn't actually make a difference in our branch, and I suspect it might break other tests, so perhaps it should be reverted. [10:43] (At the time we made the change we thought it might make a difference. It didn't) [10:44] gmb: ok, I reverted it because addError has a different signature: it wants an error, not a description as a string [10:44] Right. [11:46] hi frankban, gmb [12:11] meh [12:11] bac benji frankban gmb call in 2 [12:14] frankban, ping [13:16] Note for the record: I just pung Spamaps. Don't know if he'll reply before I have to go have my eyes poked at. [13:16] ack [13:17] bac, I'm ready to chat any time [13:17] gary_poster: ok, just a sec [13:20] gary_poster: joining hang0ut [13:21] I'm in goldenhorde. you going there or goldenhordeoneonone ? [13:22] gary_poster: i'm being forced to reinstall the plug-in. don't know why. will go to goldenhorde when i can [13:22] k [13:29] * gmb -> lunch + eye pokage [13:35] bac, horrible connection :-/ [13:35] gary_poster: that's ok. i understand what needs to be done. [13:36] bac, cool. fwiw, example: http://pastebin.ubuntu.com/903889/ [13:36] that's from current production buildbot [13:36] gary_poster: cool, thanks. [13:36] welcome [13:36] is that kill-test-pids no longer needed? i can remove it from the tree. [13:41] bac, it is not, thanks [13:46] benji, re https://bugs.launchpad.net/ubuntu/+source/lxc/+bug/959352 , the kernel team has asked for the output of apport-collect in the last comment. I tried to do it for you but it complains that I'm not the one who filed the bug. To proceed, you should have the newest updates (esp. the newest kernel) and then verify that you still have the problem (you do :-) but you can run "sudo lxc-start-ephemeral -o lpdev [13:46] -- cat /proc/self/maps" and look for "/rootfs/..." easily enough) and then run "apport-collect 959352" as requested. Once you've done so, please ping me and I'll mention it to Serge. I'll go ahead and ping him about the bug now. [13:46] <_mup_> Bug #959352: Ephemeral containers have "/rootfs" prefix in /proc/self/maps entries < https://launchpad.net/bugs/959352 > [13:47] please do that...soonish. Today if possible, and earlier is better than later. [13:47] gary_poster: ok, doing it now [13:47] thank you [13:54] why do the unity devs hate me? [13:56] It's because you run on a Mac. Wait, no... [14:04] heh [14:09] * benji reboots. [14:16] gary_poster: done [14:45] arg! I suspect my slave problems yesterday (well, the important ones at least) were already fixed and my branch was just out of date. [14:47] benji :-/ [15:16] frankban, hey. Small point, but I just converted all of your lpsetup cards from "Miscellaneous" to "Improvement". As a rule, Miscellaneous cards should only go in the Miscellaneous lanes. The only exception I'm OK with is when a slack time task does not result in any visible change, such as "study Go". Most slack time jobs will be "Improvements" I suspect. No big deal, but it helps analysis of the kanban boa [15:16] rd. [15:17] gary_poster: yes i've seen you updated the cards, thank you [15:19] gary_poster: I am seeing other cls.fixture.cleanUp errors :-/ [15:25] frankban, :-/ [15:25] frankban, do you wanna confer again, or dig some more? [15:27] gary_poster: I have an idea, and a question, so, if you have a minute... [15:27] sure frankban . going to goldenhorde [15:39] http://bazaar.launchpad.net/~gary/zope.testing/3.9.4-p4/revision/7#src/zope/testing/testrunner/runner.py [15:49] gary_poster: SpamapS: gmb: its been pre-empted by getting more urgent stuff into the distro (since we can ship charm-tools in the PPA) SpamapS: gmb: (re the python-shelltoolbox) [15:49] gary_poster: Clint says he hopes [15:49] to get it done soon. [16:34] gmb, ack, thanks for heads up [17:44] benji, you want any pairing, or are you good to go? ON a related note, should I remove the "Slave charm is generating errors" card, or is that still your focus? [17:45] gary_poster: I don't think I'm in a particularly good pairing position right this moment. [17:46] the slave charm is still generating errors, just validated the --use-html fix (there were actually two places we needed it, one had it and one didn't) [17:46] benji, cool. odd that everything was working for me, but perhaps this bitrot somewhere. bac, what about you? I'm about to go play with Go 1 on slack time if noone needs me :-) [17:47] * bac reads [17:47] gary_poster: nah, i don't think it would be useful atm [17:47] Go Go! [17:52] bac, :-) cool [18:30] status: I'm trying to figure out why the ssh in the build step results in a "Permission denied, please try again" (http://ec2-23-20-227-121.compute-1.amazonaws.com:8010/builders/lucid_lp/builds/1/steps/shell_6/logs/stdio) [19:35] gary_poster: did you get my email a while back about switchpointideas.com? [19:55] gary_poster: any idea how the slave could get into the state where this happens? https://pastebin.canonical.com/63291/ [19:55] (I can provide more detail if it would help.) [20:02] oh! I have an idea... [20:13] bac, benji, sorry, my notification story has reachd a new low [20:14] heh [20:14] bac, no, I don't think so. Just did a search [20:15] yeah, did another search. Please send again bac [20:15] gary_poster: ok [20:16] gary_poster: basically i was asking if you thought it was an appropriate use of a conference day [20:17] benji, that *might* happen if the lptests container were created by the root user, with its home directory bound.. [20:17] as opposed to buildbot's [20:17] gary_poster: yeah, that's what I was thinking. However I can't figure out how that would have happened. [20:17] benji, I assume buildbot's dir is not bound? [20:18] gary_poster: what do you mean by "bound"? mapped in the container? [20:18] if so, it is not (/var/lib/buildbot is empty) [20:24] bac, "change levers" in first para. uh-oh ;-) sounds really cool. Ubuntu fits in with the idea, certainly...I'll run it past flacoste, but it has my +1 [20:25] benji, yes [20:25] gary_poster: cool. wanna go? [20:25] you can also see in fstab of container [20:26] bac, does sound pretty cool :-) yeah, that would be fun to go to. Will check out schedule, but looks good so far [20:27] gary_poster: it may be a bit soft for a conference but something i'm personally interested in. and the venue is *amazing* [20:27] so i'll probably go if flacoste blesses. would be fun if you came along. [20:29] just asked, we'll see [20:29] benji, so pretty clearly that's the problem [20:30] is this a fresh instance since you got the updated charms? [20:30] I am tempted to try myself [20:30] gary_poster: i'm confused -- where does setuplxc.py live now? [20:30] my ec2 bill is > $300 this month :-P [20:30] bac, launchad tree, utilities dir [20:31] gary_poster: so it is. i looked but didn't see it there [20:32] gary_poster: yow! [20:32] * benji goes to check his. [20:32] :-) [20:32] I had three giganto instances running for a couple of days, running experiments, and I was also the one getting things working, so I'd expect mine to be higher than most other folks [20:33] mine is $139.61, more than it's been but not too bad [20:33] bac, flacoste sez that if we make a pitch to him about what we expect out of the conference, he'll be cool with it. I'll look into it later and maybe we can confer [20:34] great [20:36] so benji, setuplxc is the issue, somewhere, not the charm. If you still have the problem tomorrow, I will dupe and pair, 'cause I am pretty sure I have the most experience with seeing this thing actually work, at least lately [20:45] benji, I'll also be curious to see what changes, if any, have gone into our charm and setuplxc since it worked for me. I still have the version of the charm I used, AFAIK. I think. :-P [21:07] gary_poster: wow, looks what I just read: http://www.opticality.com/blog/2012/03/28/starting-a-new-life-post-zope-corporation/ [21:08] wow benji