[01:09] mwhudson: did you manage to check my fix for https://bugs.launchpad.net/maas/+bug/1307779 [01:09] ? [01:09] Ubuntu bug 1307779 in MAAS "fallback from specific to generic subarch broken" [Critical,Fix committed] [01:09] bigjools: no [01:10] i've more or less given up on armv7 :/ [01:10] mwhudson: well hopefully if it doesn't work now I can blame the simplestreams data :) [01:10] ah [01:10] bigjools: do you want me to try? i can probably dig up the relevant details [01:10] well if it's important to you, sure [01:10] but not a big deal, I already checked the use case I was fixing [01:11] ok [01:11] i'll leave it for now then :) [01:16] no worries === vladk|offline is now known as vladk === CyberJacob|Away is now known as CyberJacob === CyberJacob is now known as CyberJacob|Away [08:11] allenap: arg, your new lease parser didn't make it into 1.5.1 :/ [08:12] allenap: I think it would have made the DHCP bug less severe. [08:23] rvba: How very annoying :-/ [08:26] allenap: If we get good feedback from the field about that fix, I think it's worth pushing for a 1.5.2 release. The proper fix for the DHCP problem is going to take time and if the new parser alleviates the problem (and I think it does), it's worth releasing it. [08:53] rvba: Is that something we talk to lutostag about? [08:54] allenap: yeah, probably. [08:54] Anyone up for a tiny review? https://code.launchpad.net/~rvb/maas/packaging.trusty-upstream-rev/+merge/220765 [08:59] I'll take it. [08:59] Being nasty to someone might cheer me up. [09:04] Aah, to https://bugs.launchpad.net/maas/+bug/1322336 is basically because we do the keyring writing in the wrong place… damn. I’ll get on that now. [09:04] Ubuntu bug 1322336 in MAAS "import_boot_images crashes with KeyError on 'keyring'" [Critical,In progress] [09:06] allenap, remember that occasional lock-related startup failure? Any opinions on my guess as to the cause, i.e. ordering of decorators? [09:07] jtv: I need to reread your analysis. Sorry I haven’t replied sooner. [09:12] allenap: I can sum it up very briefly right now if that makes it easier to digest. [09:13] jtv: Sure :) [09:13] "One decorator makes the function synchronous, another makes it grab a lock; but the ordering suggests that this means 1. grab lock, then 2. run synchronously." [09:13] And so maybe the reactor goes on to run other tests during 1. [09:14] Or wait, I mis-stated that. [09:14] Maybe the reactor _already was_ running another test. [09:14] Or maybe it just arbitrarily chooses to yield sometimes even though the lock is available. [09:15] And so the test continues without actually having run that 'synchronous' function. [09:17] jtv: I think the already-running-another-test hypothesis is the most likely of those. [09:17] Either way, I guess we just can't expect the _call to_ that function to run synchronously even if we do have a right to expect _the function_ to run synchronously. === vladk is now known as vladk|offline [10:25] jtv: thanks for the review. Sorry that branch was so tiny; you didn't have much room to unleash your reviewing wrath ;). [10:26] Yeah. I'm all pent-up now. [10:32] gmb: meanwhile, if you put your branch up for review, I'm happy to have a look at it while the test is running in the lab. [10:32] rvba: Thanks. I’ll have the MP up presently. [10:32] Cool. [10:35] rvba: https://code.launchpad.net/~gmb/maas/bug-1322336/+merge/220783 [10:35] gmb: on it [10:35] Thanks. [10:39] jtv: Did you get a response from Alexis re: LXC? [10:40] gmb: she forwarded, and there was one reply saying "that's easy innit," and I said I'd like to know from a real-life example of a problematic setup. Nothing more. [10:42] Ha. [10:42] jtv: what's stopping us from deploying stuff on canonistack (with LXC) and testing the firewall ourselves? [10:43] rvba: only the risk of not reproducing the real problem setup, as far as I know. === vladk|offline is now known as vladk [10:57] rvba: Thanks for the review. [10:57] np [10:57] * gmb -> out for an early lunch. === vladk is now known as vladk|offline === vladk|offline is now known as vladk === marlinc_ is now known as marlinc [13:01] jtv, allenap, rvba: A review for someone, if you have a sec… all code-removal: https://code.launchpad.net/~gmb/maas/bug-1322606/+merge/220800 [13:01] I'll take it. [13:02] Ah, lovely conflicts coming. [13:02] But I'll manage. [13:03] Done. === melmoth_ is now known as melmoth [14:01] gmb: did you figure out what the problem was? With the import task I mean. === vladk is now known as vladk|offline === roadmr is now known as roadmr_afk === roadmr_afk is now known as roadmr === vladk|offline is now known as vladk === CyberJacob|Away is now known as CyberJacob === vladk is now known as vladk|offline [19:49] does MAAS support aggregating NICs for using LACP, and VLAN tagging yet? [19:49] i remember reading it in the blueprint last year sometime [23:44] How we can define more than 2 DNS server (IP) in MAAS? | http://askubuntu.com/q/471302 === CyberJacob is now known as CyberJacob|Away