[11:53] * gmb uploads another version of charm-tools, goes for lunch. [13:08] benji frankban gmb call in 2 [13:10] frankban, ping [13:39] gary_poster: I no longer think this landing error is a local config problem [13:39] benji, ok [13:39] I'm going to ask on -dev to see if anyone has seen it [13:40] ok benji. this is the same unicode thing as before, or different? [13:41] gary_poster: nope: https://pastebin.canonical.com/61922/ [13:41] yow, those colors are garish [13:41] benji, ugh, no idea [13:48] frankban, we've been asked to consider using lp-dev-utils [13:48] instead of a new project [13:49] lp-dev-utils has ec2 [13:49] do you have any pros or cons to that? [13:50] gary_poster: I've seen, robert. [13:51] frankban, yes, and flacoste has mentioned as well [13:56] gary_poster: I am taking a look at lp-dev-utils, no problem in landing an lpsetup package in there. I think then you can create a deb recipe to install only that package, depending on python-shell-toolbox, but I'm not sure [13:57] gary_poster: it seems that I have comitted the sin of running a VM on oneiric [13:58] gary_poster: you offered to land this branch for me, I suspect that's the sanest route; I'll then work on the MAC address issue and then try to figure out when would be the best time to try building a precise VM again [13:58] gmb: did you ever succeed in getting your precise VM up? [13:58] benji, Yes, it's what I'm working on at the moment. [13:58] so there is hope [13:59] I can try and land your branch for you if you'd like, though I think my LP setup is woefully out of date atm... [14:04] frankban, cool [14:04] benji, will do [14:04] gary_poster: thanks [14:05] I'll ping soon; filing bugs right now [14:06] benji, do you have a version of lxc-start-ephemeral that is, or demonstrates, what we need? I'd like to attach the script, and ideally a diff of the script against the current version, to the bug I'm filing. [14:06] pastebin would be good, or even better might be an MP with a branch that has the change [14:09] benji, lxc bug is https://bugs.launchpad.net/ubuntu/+source/lxc/+bug/949956 [14:09] <_mup_> Bug #949956: lxc-start-ephemeral keeps same MAC address, and so same IP address, as base instance < https://launchpad.net/bugs/949956 > [14:10] gary_poster: cool; I suppose we should go the same route as we have with the other lxc-start-ephemeral improvements and submit a branch [14:12] benji +1 [14:12] benji, please remind me of the lp branch I need to land [14:13] gary_poster: bzr+ssh://bazaar.launchpad.net/~benji/launchpad/more-integration [14:18] whee [14:19] sigh [14:19] thanks benji, on it [14:29] frankban, call in 1 or so [14:29] https://talkgadget.google.com/hangouts/extras/canonical.com/goldenhordeoneonone [14:30] benji, branch is playing in pqm [14:30] yay! [14:30] It appears to prefer the game of pinochle [14:31] benji, landed [14:31] awesome [15:29] gmb, https://talkgadget.google.com/hangouts/extras/canonical.com/goldenhordeoneonone whenever you wish [15:29] gary_poster, Okay. Firing up firefox now. [15:29] cool [16:41] has anyone seen this from lxc-start? lxc-start: invalid sequence number 1. expected 2 [16:42] gary_poster: I am having problems landing my setuplxc branch, http://pastebin.ubuntu.com/874789/ [16:42] gary_poster: I am having problems landing my setuplxc branch, http://pastebin.ubuntu.com/874789/ [16:43] sorry for double post [16:43] frankban: I was having a different problem caused by running a bzr-pqm that was too new for my bzr (because I was on oneiric) [16:44] frankban, never seen it [16:44] benji, never seen that either :-/ [16:45] benji: hum, I am using a lucid container to run lp-land [16:45] frankban, that may well be it. I did get bzr lp-land to work [16:45] in lucid container [16:45] without any issue [16:46] of course, I haven't updated the container lately [16:46] gary_poster: i will retry after and apt upgrade [16:46] frankban, ok [16:50] I'm going to try an update and reboot. If I don't come back it's because my laptop exploded and I'm dead now. [16:51] benji, no more plastic explosive attached to your hard drive connections please [16:52] don't tell me, they keep putting C4 in my apt-get upgrade! [16:52] heh [17:05] mmm, gary_poster, what is the default location for lp-land? I've used bzr+ssh://bazaar.launchpad.net/~launchpad-pqm/launchpad/devel, and maybe that's the error and I should use lp:launchpad [17:08] * benji reboots [17:15] evand's judgement on our packaging woes: "I suspect most of your problems are that it's a hideous mess of overlapping, far too flexible technologies" [17:16] That doesn't help us, but it sure as hell made me feel a bit less HULK CRUSH. [17:32] gmb, lol, and :-( [17:32] frankban, aren't they one and the same? [17:33] gary_poster: I've spotted the problem, and the branch is landed [17:33] frankban, great. so lp-land is fine in lucid after all? [17:34] gary_poster: yes, the problem was that I didn't know that the location you can provide to lp-land is the MP location, not the target branch. [17:35] frankban, ah! right [17:35] I've made another MP with the correct target branch [17:35] I saw, yeah, and merged it, cool [17:36] * gary_poster returns to a bit of lunch break [17:43] * gmb -> evening things [18:52] benji, if you'd like to pair we could do that whenever it is convenient for you; otherwise I'm staring at a parallel-testing-related bug (609986) and we can talk at 2:30 [18:54] gary_poster: I'm starting to look at the test failures in http://ec2-50-16-2-233.compute-1.amazonaws.com:8010/builders/lucid_lp/builds/0/steps/shell_8/logs/stdio [18:55] benji, I don't see an ssh error in that, yeah? [18:56] gary_poster: right, none in sight; the MAC hack seems to have worked [18:57] benji, great, I'll update the board accordingly. That's a weird result there, though...we don't usually see successful test results when running in parallel [18:57] benji, and why do we have a kill at the end--is one parallel process done? [18:58] gary_poster: right, one must have finished; I assumed that's why we have the successes too; are successes reported at the end? [18:59] benji, I don't *think* so; also, if you look forte first "successful, it is a BaseLayer tearDown [19:00] forte -> for the [19:01] benji, "lxc-wait: bind : Address already in use" is also concerning [19:01] I'm not surprised your typos are musical terms. [19:01] :-) [19:02] I /think/ that error is OK-ish; it's the stopping script trying to do an lxc-stop. I do think there's a bug in lxc though that we may have to track down (otherwise we'll have tons of left-over ephemeral containers lying around. [19:03] or maybe it's the lxc-wait, one or the other [19:03] yeah, dunno [19:03] I'm in https://talkgadget.google.com/hangouts/extras/canonical.com/goldenhorde but also listening to music, so you can regard that as an invitation that you do not need to accept :-) [19:05] benji, I'm also curious that there has been no output since the kill. Have you gotten on the machine yet to see what the story is? [19:06] gary_poster: poking around now, I can verify that there is one still running (and using CPU) [19:06] huh [19:07] although, I'm a little surprised to see that it shows up as /usr/bin/python2.6, I would have expected it to be hidden inside something lxc [19:12] I'm not sure how this kind of virtualization looks like to the host [19:12] This is concerning too" [19:12] OperationalError: source database "launchpad_ftest_template" is being accessed by other users [19:12] DETAIL: There are 1 other session(s) using the database. [19:17] hmm, the remaining bin/test process looks like it's stuck in a select loop waiting for something to happen [19:20] :-/ [19:30] benji, come on by https://talkgadget.google.com/hangouts/extras/canonical.com/goldenhorde when you are ready [19:31] * benji fires up the hordemobile. [20:27] * gary_poster switches to the other os [20:52] benji, I have collapsed in exhaustion (woke up @ 3 AM with stomach pains) but I'm in goldenhorde :-)