/srv/irclogs.ubuntu.com/2014/08/24/#juju-dev.txt

thumpermorning folks21:41
=== wallyworld_ is now known as wallyworld
thumperhi waigani22:45
thumperwallyworld: morning22:45
wallyworldhey22:45
waiganithumper: hello22:45
thumperwallyworld: https://bugs.launchpad.net/juju-core/+bug/136028622:45
mupBug #1360286: ppc64el and i386 unittests cannot find tools <ci> <regression> <juju-core:Triaged> <https://launchpad.net/bugs/1360286>22:45
thumperwallyworld: my guess is axw's bootstrap refactoring22:45
thumperwallyworld: do you think we should revert or fix?22:45
thumperwallyworld: blocking CI22:46
wallyworldcould be, i'll look if my network connection satys up long enough :-(22:46
wallyworldlet me look first22:46
waiganithumper: checking the user exists I can only see an unexported func state.checkUserExists22:46
thumperwaigani: what is this for?22:47
waiganithumper: AddEnvironmentUser22:47
waiganithumper: checking if user exists before adding it22:47
thumperwaigani: ok, there are two main possibilities22:47
thumperwaigani: 1) check at the apiserver level using state.User(name) to load the user22:48
thumper2) look inside the actual state method to add an environment user22:48
thumperI'd probably prefer 2.22:48
waiganithumper: okay, also how do we tell if they are local?22:48
thumperwaigani: the tag Provider method is "local"22:49
waiganithumper: thanks22:49
wallyworldthumper: the failing test can be removed since the logic has changed - i'll do a branch with it skipped to unblock ci and confirm with andrew and then remove it22:54
thumperwallyworld: cool22:55
thumperwallyworld: http://juju-ci.vapour.ws:8080/job/github-merge-juju/buildTimeTrend23:15
thumperwallyworld: FYI looks like Build #374 (21-Aug-2014 07:52:39) axw environ-bootstrap-upload-tools is the one that started big time changes23:15
wallyworldthumper: our build times have also been going up because tests are really unreliable and we now fail back to -p 2 most times23:22
thumperah23:22
* thumper suggests that we only check once23:23
thumperfail faster if we are going to fail23:23
wallyworldtrouble is the 2nd run normally works23:27
wallyworldbut is much slower than with full parallelism23:27
wallyworldwe just need to fix the freaking tests23:28
thumperwallyworld: agreed23:36
thumperthe replicaset test seems to be overly fragile23:37
thumperand long23:37
thumpercan we not force that out of our normal test run into something more suitable?23:37
thumperand write a robust test?23:37
wallyworldyou would think so23:37
wallyworldalso, jujud tests are horrible23:38
wallyworldand there's a failing uniter test23:38

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!