[01:28] smoser: recall how pylint has started complaining about paramiko and simplestreams when I moved the code up to the super class? Well it turns out it is because the platform directories do not have an __init__.py [01:28] once I add a __init__.py into the nocloudkvm directory it finds other import errors, including paramiko === shardy is now known as shardy_afk === shardy_afk is now known as shardy [14:32] horay. i'm not really sure why but [14:32] https://jenkins.ubuntu.com/server/job/cloud-init-copr-test-7/145/console [14:33] is fixed now. [17:14] smoser: 17.2 release merge proposal up for bionic https://code.launchpad.net/~chad.smith/cloud-init/+git/cloud-init/+merge/335267 [17:39] blackboxsw: gar... [17:39] tip-pylint moved, and caught what seems to be some valid things [17:39] sounds good. will run against tip to see [17:39] and will propose a fix to master [17:40] well, 2 feelings [17:40] a.) we dont just chase pylint upstream just because [17:41] b.) but valid issues are good to fix. (it seems like the W0715 are valid) [17:41] i swear that i had seen complaints like the 'logging-not-lazy' before [17:42] so we have the tip-* tox targets, when is a good time/process for us to check/validate/fix tip-* related issues? [17:42] very lazily. [17:42] maybe post SRU/release we make a concerted effort to check and fix ? [17:42] we moved them out so that we could easily run them, to be aware of things [17:43] but they should not ideally fail your c-i [17:43] I agree with that [17:43] how would there be any tip-* failures on master, given every merge goes through CI and we run tip-*? [17:43] as we didn't get a integration run because of that failure [17:43] we shouldn't continue to have tip-* in our CI standard path. [17:43] powersj: new tip-pylint update [17:43] ah missed that [17:43] this is my "dude on the internet" [17:43] heh [17:43] he broke our c-i [17:44] powersj: can we remove pylint-tip from the merge proposal runs ? [17:44] i think i'd be happy to have it run on trunk nightly in a separate job entirely [17:44] sure it can still be a part of ci-nightly [17:44] I'll pull from merge reviews [17:44] so every ci-run is racing against tip-pylint potential updates. nightly would be excellent [17:45] smoser: here is what we run today: https://paste.ubuntu.com/26189737/ [17:45] want me to remove line 4 only? [17:46] drop tip-* [17:46] they're all broken by dude on the internet [17:46] 'sh' 'tox' [17:46] that is funny [17:47] it's a jenkins pipeline syntax [17:47] really funny in that its wrong [17:47] ie, if you just read that [17:47] sh tox [17:47] should run sh on the file 'tox' [17:47] but what its doing is: sh -c tox [17:47] which is just pointless ;) [17:47] smoser: jenkins jobs updated and deployed [17:48] ok. can you re-run on blackboxsw's branch [17:48] as i was just going to ignore [17:48] but i'm happy to jsut re-visit with a happy c-i in 15 minutes [17:49] * powersj clicks rebuild [17:51] so, that aside, , we still want to fix a couple of these http://pastebin.ubuntu.com/26189753/ [17:51] anyone care if I take a pass at a trivial there on master ^ [17:51] s/care/object [18:36] https://code.launchpad.net/~chad.smith/cloud-init/+git/cloud-init/+merge/335271 pushed [18:36] errand [19:10] smoser: thx for review [19:11] I just finished changing the VPC code to generate and delete on each run, however, as you pointed out there is a limit of 5. I still think making a new VPC each time is the preferred option. [19:11] But will have to catch the exception when we hit the limit [19:11] should we hit the limit, rather [19:19] powersj: i'm sorry [19:21] powersj: what env (series) is ci running again? [19:21] system packages on xenial are too old for boto3 w/ your branch. I'm switching to artful [19:22] as boto3.Session() doesn't have a region_name attr defined [19:22] in 1.2.2 [19:22] blackboxsw: i hate .format() [19:22] its so long [19:22] "my {0}".format("value") [19:22] "my %s" % value [19:22] smoser: agreed, but it's explicit. if we want ot use %s' % everywhere I'm ok with that (though I wasn't sure if pylint was also trying to move away from iit [19:23] blackboxsw: if you use tox it will pull latest ;) [19:23] I was under the impression that .format() was 'the future' but that's maybe a misconception [19:23] or the version you need [19:23] yeah i had gotten that vibe also [19:23] but that doesn't mean i can't hate the future. [19:23] what fun would it be to be old if you couldnt hate the stuff those kids are doing [19:24] "get off my lawn" [19:24] .. crazy kigs [19:24] kids even [19:24] dont get me started on how kids are spelling these days [19:24] man [19:25] * blackboxsw sacrifices my boto3 system package to appease the tox venv 'gods' [19:26] if you're going to use .format, then it just seems wrong to not use {name} and .format(name=value) [19:26] ie, the change is in the name of "explicit" or readability and {0}, {1}, i dont think make that. [19:32] smoser: yeah, I know I also dislike all the additional characters.... format ++ the cost of named params eek [19:33] but yeah I'll tweak it. it is better to be explicit [19:33] i approved [19:33] will still fix it ;) [19:33] and grabbed your other now. the upload. [19:47] powersj: boo. [19:47] blackboxsw: ? [19:47] :/ data = self._decompressor.decompress(data, n) [19:47] MemoryError [19:47] time for me to buy a bigger computer [19:47] got that running tox citest ;) [19:47] O.o [19:48] can I get full trace? [19:48] curious what caused that [19:48] http://pastebin.ubuntu.com/26190389/ [19:49] why? "Enter passphrase for key '/home/csmith/.ssh/id_rsa': " [19:50] crazy that is only during the install [19:51] ahh just recently added a check in bashrc check for ssh-agent to ssh-add my key to cache it which is used for accessing ec2 instances [19:54] hrm 20G memory used on this machine (even after killing landscape and a number of containers) [19:54] ahh ha. [19:55] chrome Virt == 21G [19:59] firefox59 time [19:59] blackboxsw: you need to use Tab Wrangler and autoclose tabs after 24 hours ;) [20:02] powersj: does tab wrangler take more memory than chrome? :) [20:02] hah [20:03] blackboxsw: also The Great Suspender for chrome; that's been really nice [20:11] powersj: hah! too true, this instance of chrome only had 4 tabs (as I don't use my desktop much) but it had been up for > 30 days [20:11] goodness [20:36] ~30 mins powersj on your ec2 review. almost there, couple nits [20:37] blackboxsw: thanks got my list from rharper and smoser :) so getting yours will be nice [20:37] * blackboxsw wants 'coal' for xmas [20:38] blackboxsw: i spent liek 3 hours. :) [20:39] I'm only 1 hour in ;).... I'm going to be efficient (and cut-n-paste from ec2-launch) [20:39] ... and it's easy to just leverage rharper and smoser's review comments ;) [20:57] lol [20:58] due to a recent [20:58] change in SRU bug verification policy fixes now have to be marked with [20:58] per-release tags (i.e. verification-done-$RELEASE). [21:41] powersj: well a bit harder than I thought. review posted on your ec2 integration branch. great work [21:41] rharper: yeah, cloud-init srus required as much. [21:42] blackboxsw: well, I just meant; it seems strange to email that now after it's already done [21:42] I'm happy with verification-$release [21:42] blackboxsw: thanks [21:50] * blackboxsw reworks the git-publish script to not depent on cloudinit utilities [21:57] well, that was what smoser wrote [21:57] wrong channel [22:05] here comes a flood of bug comments. [22:17] blackboxsw: very confused, your email to me was on the tip-lint-fixes branch [22:17] https://code.launchpad.net/~chad.smith/cloud-init/+git/cloud-init/+merge/335271 [22:18] powersj: sorry, how did my auto-review against my own branch send you and email? [22:18] too many tabs open again? [22:18] nah was running on the commandline [22:18] well you put a comment directed to me in that review [22:18] ohhh above [22:18] bummer [22:19] hah, and yes too many tabs [22:19] shoot [22:19] did I even submit your comments on the right branch.... checking [22:20] .... and i did not [22:20] gah [22:20] * powersj is very confused [22:21] powersj: it's the person you work with, man. Fridays... [22:22] ok posted your review comments for realsies powersj [22:22] ... and yes it's friday again, and I just closed 15 tabs (still have > 15 remaining) [22:22] <-- problems [22:23] hopefully powersj you got the review comments [22:23] * blackboxsw reloads the proper MP again [22:25] blackboxsw: thank you [22:31] smoser: any objection to me adding a few log debug statements while creating/deleting VPCs, launching instances, etc.?