[09:09] <frankban> gmb: morning, ready to pair when you are
[09:09] <gmb> Hi frankban. I'll be ready in about 20 minutes time; just need to send off some emails first.
[09:09] <frankban> gmb: cool
[09:34] <gmb> frankban, Okay, I'm ready. I'm in https://plus.google.com/hangouts/_/extras/canonical.com/subunittags (without video, but I'm sure you'll survive :))
[09:37] <frankban> gmb: joining
[11:05] <bac> 2-factor authentication is pissing me off
[12:12] <gary_poster> bac benji frankban gmb, apologies.  Was trying to get the convoy packaging issue resolved, and now I'm blocked on a family issue.  Hopefully I'll be ready in 3
[12:13] <bac> ok
[12:15] <gary_poster> bac benji frankban gmb call in 1 or asap
[12:35] <gmb> Rats. Should probably have stayed in the hangout...
[12:35] <gmb> face
[12:36] <gary_poster> gmb, you come up with such amusing hangout names, so I nominate you to come up with our zope:layer hangout
[12:36] <benji> bac: you should get some tea and then we'll hang out
[12:36] <bac> benji: :)
[12:36] <bac> you know i was just about to type that
[12:36] <gary_poster> no pressure on the being amusing bit :-P
[12:41] <benji> my mind reading must be working well today
[12:41] <gmb> gary_poster, frankban: https://plus.google.com/hangouts/_/extras/canonical.com/layeruponlayerofconfusion
[12:41] <gary_poster> very nice, though I am all alone there
[12:46] <benji> bac: when the tea is brewed, the horde awaits
[12:46] <bac> benji: yes, i had to make a new pot.  i'm ready now.
[12:47] <bac> benji: normal bat channel?
[12:47] <benji> bac: yep
[13:42] <gary_poster> benji I meant to tell you on call: get MyUnity from software center and you can have a nice way to control size of workspaces, along with other stuff
[13:43] <benji> gary_poster: oh yeah! I forgot about that. I actually have it installed but forgot that it existed.
[13:43] <gary_poster> heh cool
[13:44] <gary_poster> bac & benji, hey I'm sorry to interrupt but could you join us in https://plus.google.com/hangouts/_/extras/canonical.com/layeruponlayerofconfusion ?  We have a solution, but the only explanations as to why it works don't make sense, so we'd like to see if either of you have insight.  We can constraint to 10 minutes: over at 10:55
[13:45] <gary_poster> constrain
[13:52] <benji> bac: I think I'm here but I don't see you.
[15:14] <gary_poster> bac benji frankban gmb, fwiw I wrote some annoying instructions on how to get the list of workers from the subunit links on the ResultsLog page.  The instructions are at the bottom of the page (https://dev.launchpad.net/ParallelTests/ResultsLog).  Improvements & questions welcome.
[15:16] <bac> gary_poster: benji will save us from all that soon i hope
[15:16] <gary_poster> bac, via the buildbot changes?  Yes, though we may still need/want to be able to analyze the old subunit outputs
[15:24] <frankban> gary_poster: since you already have some context, you may be interested in the lp-lxc-ip MP: https://code.launchpad.net/~frankban/lpsetup/lp-lxc-ip/+merge/103502
[15:25] <gary_poster> cool frankban
[15:26] <gary_poster> will look
[15:26] <frankban> thanks gary_poster
[15:52] <gary_poster> frankban, that branch is awesome!  you have to talk me through some of it so I'm sure I understand what is going on. :-P https://talkgadget.google.com/hangouts/_/extras/canonical.com/goldenhordereview ?
[15:52] <frankban> gary_poster: joining
[16:15] <gmb> frankban, Is lpsetup now considered a useable replacement for rocketfuel-get, or should I consider to tell people to use rf-g for now?
[16:39] <gmb> gary_poster, Do you know the answer to that ^^?
[16:45] <frankban> gmb: lpsetup is not ready to replace rocketfuel
[16:45] <gmb> frankban, Okay, cool. Thanks.
[16:59] <gary_poster> frankban, I am running sudo nosetests.  (at least) one test takes a really long time, right?  I should not assume that it is hanging?
[17:00] <frankban> gary_poster: yes, it creates an lxc, and the first time it could take some minutes
[17:02] <gary_poster> ok cool
[17:04] <gary_poster> frankban, fwiw http://pastebin.ubuntu.com/945997/
[17:04] <gary_poster> trying again
[17:05] <frankban> gary_poster: hum, do you have the lxcbr0 interface up?
[17:05] <gary_poster> yes frankban
[17:05] <gary_poster> and failure repeated
[17:07] <frankban> gary_poster: could you please print the ip before the failing lines and the run nosetests -vs?
[17:08] <gary_poster> sure
[17:10] <gary_poster> frankban, it passed now :-/
[17:10] <gary_poster> 10.0.3.6 fwiw
[17:11] <gary_poster> ah interesting frankban
[17:11] <gary_poster> frankban, I ran it again without -vs for the heck of it
[17:11] <gary_poster> I got one failure (test_get_ip_
[17:11] <gary_poster> )
[17:11] <gary_poster> and the ip was 10.0.3.182
[17:12] <frankban> gary_poster: this is another test run, right?
[17:12] <gary_poster> I did it again and it passed
[17:12] <gary_poster> then I did it again and test_restart failed
[17:12] <gary_poster> with ip 10.0.3.28
[17:12] <gary_poster> yes frankban,separate test runs
[17:14] <frankban> gary_poster: the ip is retrieved correctly, maybe this is a race condition, could you please try to add time.sleep(1) before the failing lines?
[17:15] <frankban> gary_poster: maybe the dns needs a second to register that ip, and maybe I should just test that the ip is actually an ip, rather then using socket.gethostbyaddr
[17:16] <gary_poster> ok
[17:18] <gary_poster> frankban, with time.sleep(1) I got both failures :-P
[17:18] <gary_poster> 10.0.3.38 in both cases
[17:19] <gary_poster> I'll increase sleep for the heck of it
[17:19] <gary_poster> 1 full pass with 5 seconds
[17:21] <gary_poster> another one...
[17:22] <gary_poster> frankban, I entered my review on the MP.  It has a new thing or two, so take a glance if you would.
[17:22] <gary_poster> another full pass...
[17:22] <gary_poster> another one...
[17:23] <gary_poster> another...
[17:23] <frankban> thanks for the review gary_poster. so, it seems a race condition
[17:23] <gary_poster> frankban, time.sleep(5) seems to do it, yeah.
[17:23] <benji> gary_poster: our buildbot run was bit by the _buffered_calls AttributeError
[17:23] <gary_poster> (another one)
[17:24] <gary_poster> benji, really!  super annoying
[17:24] <frankban> gary_poster: in that case I have two choices:
[17:24] <benji> gary_poster: yeah, we're going to have to fix that before we can continue
[17:25] <gary_poster> ok benji.  good luck.  lemme know if you'd like another pair of eyes.  I cc'd you on the email to jml about it, yeah?
[17:25] <gary_poster> no reply
[17:25] <frankban> 1. generalize the retry decorator and use it with socket.gethostbyaddr
[17:25] <benji> yep
[17:25] <frankban> 2. just test the the returned value is a valid ip
[17:25] <gary_poster> 3. (potentially painful) figure out why there is a race condition and what you can wait for instead
[17:26] <gary_poster> I kind of prefer 1 to 2 frankban, but will defer
[17:27] <frankban> gary_poster: i prefer 1 too, for 3, I have to reproduce the race condition (never seen before) in my machine
[17:29] <gary_poster> frankban, sure.  go for #1
[17:31] <bac> hi benji
[17:34] <benji> bac: take a look at this http://ec2-23-20-226-202.compute-1.amazonaws.com:8010/builders/lucid_lp/builds/0/steps/shell_9/logs/stdio
[17:34]  * bac looks
[17:36] <bac> benji: you've encountered this problem before?
[17:37] <benji> yep, we reported to jml but he hasn't had time to look at it; I guess it's our turn now
[17:37] <benji> I'm filing a real bug for it now.
[17:41] <frankban> gary_poster: I'be just pushed a revision that uses retry for those tests
[18:04] <gary_poster> frankban, passed fully three times in a row.  +1
[18:05] <bac> benji: i'm doing a review...ping me when ready
[18:06] <benji> bac: ok
[18:18] <benji> bac: https://talkgadget.google.com/hangouts/_/extras/canonical.com/goldenhorde when you're ready
[18:18] <bac> benji: ok, give me a bit
[18:47] <bac> benji: joining
[18:47] <benji> k
[19:52] <gary_poster> benji, what is the bug number of the buffer attribute problem you filed for testrepository?  (I'm assuming you are working on it, and will update the board as such)
[19:55] <gary_poster> bug 988481
[19:55] <_mup_> Bug #988481: Testr tests generate an AttributeError when run with latest testtools <Testrepository:In Progress by yellow> < https://launchpad.net/bugs/988481 >
[19:55] <benji> gary_poster: yep, that's it
[19:55] <gary_poster> thanks
[19:55] <benji> it's melting our brains at the moment, but when life gives you melted brains, you make brain soup
[19:59] <gary_poster> heh
[21:05] <gary_poster> benji and bac are listed as Ubuntu contributors for 12.04!
[21:05] <benji> heh
[21:05] <gary_poster> So I hear from sinzui
[21:05] <bac> for why?
[21:06] <benji> lxc?
[21:06] <gary_poster> lxc maybe?
[21:06] <bac> juju?
[21:06] <gary_poster> maybe
[21:07] <benji> gary_poster: I'm listening to the song listed under "Latest Single" on the lower right of  http://perhapscontraption.co.uk/ and it made me think of a group that you like that I can't remember the name of :)
[21:07] <gary_poster> :-) on call, will check after
[21:08] <bac> benji: type 'yes'
[21:08] <benji> heh
[21:08] <benji> thanks bac
[21:08] <bac> np
[21:08] <bac> i felt so helpless
[21:09] <benji> lol