| *** baoli has joined #openstack-sprint | 01:14 | |
| *** rfolco has quit IRC | 02:25 | |
| *** baoli has quit IRC | 03:35 | |
| *** hieulq has quit IRC | 06:30 | |
| *** hieulq has joined #openstack-sprint | 06:33 | |
| -openstackstatus- NOTICE: Precise tests on OSIC provider are currently failing, please stop your checks until the issue is resolved. | 11:17 | |
| -openstackstatus- NOTICE: DSVM jobs on OSIC currently failing because of IP collisions, fix is in the gate - https://review.openstack.org/#/c/357764/ - please hold rechecks until merged | 11:22 | |
| *** SotK has quit IRC | 11:26 | |
| *** waynr has quit IRC | 11:26 | |
| *** SotK has joined #openstack-sprint | 11:27 | |
| *** waynr has joined #openstack-sprint | 11:27 | |
| *** cdelatte has joined #openstack-sprint | 11:42 | |
| *** baoli has joined #openstack-sprint | 11:42 | |
| -openstackstatus- NOTICE: OSIC has burned through the problematic IP range with failures, things should be back to normal now. | 11:50 | |
| *** rfolco has joined #openstack-sprint | 12:10 | |
| *** baoli has quit IRC | 12:51 | |
| *** baoli has joined #openstack-sprint | 13:03 | |
| *** baoli has quit IRC | 13:34 | |
| *** baoli has joined #openstack-sprint | 13:47 | |
| waynr | zaro zxiiro yo i woke up a bit late today so i probably won't be attentive here for about an hour yet FYI | 14:02 |
|---|---|---|
| zxiiro | waynr: cool, I rebased a few patches last night that I think we can easily merge. There's still the 2 patches that electrofelix brought up last time that we need to work through his comments on | 14:04 |
| *** electrofelix has joined #openstack-sprint | 14:19 | |
| electrofelix | aloha! | 14:19 |
| zxiiro | electrofelix: yo! | 14:20 |
| zxiiro | electrofelix: i rebased a few patches yesterday and added it to the etherpad that i think we can merge today | 14:20 |
| zxiiro | electrofelix: there's still the 2 ones you had comments about that we should figure out | 14:20 |
| electrofelix | zxiiro: thanks, saw that earlier when sending out the reminder email | 14:20 |
| zxiiro | ah i should have checked my email | 14:21 |
| electrofelix | ok, today I'm going to focus on "Add static methods to classes to create from config object to retain full __init__ API signature while simplifying default creation" after I've finished reviewing those 5 patches | 14:33 |
| electrofelix | as well as seeing if a test can be added to capture the issue we hit with CLI accidentally deleting jobs, I think it's a CLI issue that we failed to test the options correctly rather than an API issue. | 14:34 |
| zxiiro | I want to rebase the views patch but I suspect we should wait for a few more of these api patches merged first | 14:38 |
| waynr | okay i'm back online | 14:56 |
| electrofelix | waynr: do you want to give the patches zxiiro rebased the once over, I think they are mostly the same I'm just going through them | 14:57 |
| waynr | electrofelix: are these all the patches listed under "Patch list for landing"? | 14:58 |
| electrofelix | looks like the main change is just moving the two patches that change class hierarchy/inheritance model for builder.Jenkins and builder.Builder are moved out to be later so we can delay reviewing them | 14:58 |
| electrofelix | waynr: yeap | 14:58 |
| waynr | oh, cool | 15:00 |
| zxiiro | yeah if we can't land those 2 patches we should rebase them so that they are at the end because I think the next set of patches are pretty trivial too | 15:00 |
| waynr | yeah i'm totally down with that, the class hierarchy/inheritance model thing was more of an experiment/afterthought | 15:01 |
| waynr | by the way i used JJB 2 to great effect this past week to automatically generate 150+ JIRA tickets for a mass Jenkins migration effort we are undergoing soon, made a ticket for each job-group, job-template, and !include script matching a particular regular expression (since the migration work will involve making all these things compatible with a new jenkins plugin we are writing) | 15:06 |
| zxiiro | waynr: neat, I've been using what's in master for my daily usage and it been working for what I've been doing too | 15:07 |
| waynr | while doing this i had some ideas for improving the traceability of different JJB YAML elements by having the custom yaml.Loader in local_yaml return custom OrderedDict objects that contain metadata references to the files each dict was loaded from | 15:08 |
| waynr | (maybe even line numbers to indicate where each specific job-template, job-group, project, default, etc came from) | 15:09 |
| zxiiro | that would be great, we often have confusion from our users of where a yaml syntax error | 15:18 |
| zxiiro | is | 15:18 |
| *** larainema has quit IRC | 15:21 | |
| *** larainema has joined #openstack-sprint | 15:23 | |
| electrofelix | waynr zxiiro: is there any chance we might see usage of python<3.2 ever? I'm presuming not, can't think of anything that I can recall that would suggest it is used anyway? | 15:35 |
| electrofelix | s/anyway/anywhere/ | 15:35 |
| zxiiro | electrofelix: I don't think so. I think we want to support at least whatever comes with the distros by default. I believe these days it's 3.4 | 15:35 |
| electrofelix | just spotted that our PY2/PY3 compat uses read_file for a configparser object which only appears in py3.2 | 15:36 |
| electrofelix | zxiiro: that was my thought too | 15:36 |
| waynr | well python 2.7 is the default python on current versions of debian | 15:36 |
| zxiiro | yeah it's usually 2.7 default and 3.4 as an option | 15:36 |
| electrofelix | waynr: meant that where python 3 is being supported can you recall anything supporting python < 3.2? | 15:37 |
| waynr | nope | 15:37 |
| electrofelix | ok, cause that means we can also just switch to using read_string for the DEFAULT_CONF as well ;-) | 15:42 |
| electrofelix | waynr zxiiro: spotted a few things with https://review.openstack.org/#/c/319624/13, do we want to TODO them or fix the commit? | 15:42 |
| electrofelix | also looks like we've all missed that __all__ should be at the top of the file with https://review.openstack.org/319625, I'd say we just forgot about it | 15:46 |
| waynr | whoops, i'll fix that one up | 15:51 |
| zxiiro | waynr: electrofelix: I just noticed, i think the cli doesn't initiate the logger. So a command like "jenkins-jobs delete some_job" won't return any feedback that jobs were removed | 15:53 |
| zxiiro | if we're purposely not starting the console logger how should we handle feedback? | 15:54 |
| waynr | i don't know if we're purposely not starting it | 15:55 |
| electrofelix | zxiiro: I suspect it accidentally got removed in some of the refactoring, I recall making a change so that we weren't enabling in the API by default, as it should only be via entry | 15:55 |
| electrofelix | found it, missed it being removed: https://review.openstack.org/#/c/319610/ should have included adding a corresponding line to call logging.basicConfig(level=logging.INFO) in /jenkins_jobs/cli/entry.py | 15:57 |
| waynr | i have to go to a company all-hands (they are bribing us to attend with tasty cold brew coffee) | 15:59 |
| electrofelix | we should probably add it to the entry.py and replace the line "logger = logging.getLogger()" inside the __init__ method with it, since the file already has that just after the imports | 15:59 |
| zxiiro | electrofelix: ah ok cool | 16:10 |
| *** sebian has quit IRC | 16:14 | |
| electrofelix | zxiiro: I wonder, add logging.basicConfig() to top of file to get any logged warnings from argparser and JJBConfig, or move JJBConfig below the initial options parsing and start logging from after the initial arg parsing? | 16:15 |
| electrofelix | I'm thinking put it at the start | 16:15 |
| zxiiro | electrofelix: I always put my logger at the very top | 16:16 |
| zxiiro | more logs is always a good thing | 16:16 |
| electrofelix | zxiiro waynr: https://review.openstack.org/357994 should fix the logging | 16:18 |
| zxiiro | electrofelix: I also did the for loop in the delete todo you had https://review.openstack.org/#/c/357990/1 | 16:18 |
| *** sbadia has joined #openstack-sprint | 16:18 | |
| electrofelix | zxiiro: I'll think I might just review the tests for that before I approve based on our last experience :p | 16:20 |
| zxiiro | haha sounds good | 16:20 |
| zxiiro | electrofelix: I can confirm your patch works | 16:22 |
| zxiiro | electrofelix: What happens if i workflow+1 before verified+1 returns? | 16:23 |
| zxiiro | electrofelix: does it still go through the proper tests or skip it? | 16:23 |
| * waynr works on 319625 | 16:24 | |
| waynr | electrofelix zxiiro unless either of you have already begun I'll also address the comments on https://review.openstack.org/#/c/319624/ | 16:38 |
| zxiiro | waynr: I haven't touched it | 16:39 |
| electrofelix | neither had I | 16:40 |
| *** krotscheck is now known as the_other_clarkb | 16:42 | |
| *** the_other_clarkb is now known as krotscheck | 16:43 | |
| waynr | holy shit i hate balancing quote marks between multiple lines of the "same" string between parens | 16:53 |
| clarkb | pep8 doesn't require that fwiw | 16:54 |
| waynr | almost enough to make me want to go into farming ;) | 16:54 |
| clarkb | it accepts either balanced or extra indent | 16:54 |
| waynr | okay 319624 and 319625 are updated and ready for re-review | 16:56 |
| waynr | rereview | 16:56 |
| waynr | rearview | 16:56 |
| waynr | clarkb: hmm are we talking about the same thing? I am referring to wrapping long strings using multiple single-quoted strings and python's implied line continuation inside parenthesis | 16:59 |
| clarkb | ya, I thought you were saying that balancing indentation of that is a pain. Eg if the functiona name changes everything has to move | 17:01 |
| waynr | the multiple single-quoted strings inside parenthesis bothers me because it's annoying to change the string contents then have to change all the quote characters to match line length limits, although i imagine there must be an emacs package with a function to automate this so i'm really just complaining uselessly | 17:03 |
| waynr | electrofelix: so with regard to https://review.openstack.org/#/c/319751/ and https://review.openstack.org/#/c/319754/, I agree that we probably shouldn't inhereit from the python-jenkins module's Jenkins class but it's not clear to me if you are in favor of merging builder.Jenkins and builder.Builder...? | 17:08 |
| waynr | oh wait, actually I think I see comments indicating you are in favor of it but naming the combined class "JenkinsManager" instead, does that sound right? | 17:08 |
| waynr | if so I will start working on that now | 17:08 |
| electrofelix | I'm happy enough about merging them, if you think some of the methods from builder.Jenkins should be really be part of the jenkins api I think we should look at getting them added | 17:08 |
| electrofelix | yeap, thought a rename of the class might help distinguish that it's not the same ;) | 17:09 |
| waynr | i'll rethink whether i think stuff from builder.Jenkins should be part of the upstream jenkins API while working on it | 17:10 |
| waynr | argh...gerrit/git-review seems to make it easy to get lost in a large series of patchs like what we're working on | 17:11 |
| electrofelix | zxiiro waynr: I've done an initial stab a having both the explicit and simple API available for the Builder class https://review.openstack.org/358019 set it -1 WIP for now | 17:11 |
| electrofelix | waynr: in reusing the topic name for the local branch? | 17:12 |
| electrofelix | that's definitely one thing I'd like to look at turning off | 17:12 |
| waynr | well i'm lost right now because in this patch [https://review.openstack.org/#/c/319755/11] I moved CacheStorage out of the builder module but in this patch [https://review.openstack.org/#/c/333780/10..11] it seems to have been added back in and I can no longer see the relationship between these two in the "Related Changes" pane on the latter commit | 17:13 |
| electrofelix | waynr: I think it's an ordering thing, the rename parallelize is no longer directly in line from the CacheStorage patch | 17:15 |
| electrofelix | I think it's in order to allow it to be merged without needing to wait on the Jenkins/Builder classes changes, I guess that's going to cause some conflicts somewhere though | 17:16 |
| waynr | ...oh yeah | 17:16 |
| electrofelix | perhaps we can get zxiiro to help complete the remaining reviews and then straighten up the patch series? | 17:17 |
| electrofelix | I've +2'ed the next three patches listed in the etherpad, and the next one after that is my own so can't review it ;-) | 17:17 |
| zxiiro | I can look soon. Just finishing up some unexpected work | 17:17 |
| waynr | I +2'd your patch electrofelix ... zxiiro did you change your mind on the workflow +1 for https://review.openstack.org/#/c/357994/ for a specific reason? | 17:18 |
| electrofelix | waiting for CI to get back first? | 17:18 |
| waynr | oh yeah | 17:19 |
| zxiiro | yeah i was waiting for jenkins | 17:19 |
| zxiiro | i'm not entirely sure what happens when you workflow+1 before jenkins returns | 17:20 |
| electrofelix | ok, I need to head, I'll check again in about 2 hours, as I've a class to go to in about 40 minutes | 17:20 |
| waynr | electrofelix: cool, thanks again for leading this sprint! | 17:20 |
| zxiiro | electrofelix: cool thanks | 17:21 |
| electrofelix | zxiiro: I think it just moves it into the gate automatically, but since we also might want to check any warning on the larger set of jobs | 17:21 |
| electrofelix | it's generally better to wait until it gets back, since the warning won't stop a change from landing | 17:21 |
| clarkb | zxiiro: if the jenkins run fails its a noop, if the jenkins run passes it will enqueue into the gate and potentially merge | 17:21 |
| electrofelix | however in this case it shouldn't affect the job config comparison check | 17:21 |
| zxiiro | electrofelix: waynr ok all patches reviewed, Just waiting for Jenkisn to return on them before we merge | 17:23 |
| waynr | cool | 17:23 |
| electrofelix | ok, chat to you later or on Monday, enjoy the weekend | 17:24 |
| *** electrofelix has quit IRC | 17:24 | |
| *** baoli has quit IRC | 17:29 | |
| *** baoli has joined #openstack-sprint | 17:30 | |
| *** baoli has quit IRC | 17:40 | |
| *** baoli has joined #openstack-sprint | 17:46 | |
| *** baoli has quit IRC | 17:56 | |
| *** baoli has joined #openstack-sprint | 17:56 | |
| *** baoli has quit IRC | 18:21 | |
| *** baoli has joined #openstack-sprint | 18:26 | |
| *** baoli has quit IRC | 18:33 | |
| *** baoli has joined #openstack-sprint | 18:36 | |
| waynr | gerrit-- | 18:44 |
| waynr | or maybe i'm just stupid | 18:44 |
| waynr | two related sets of patchsets diverged in a yellow wood | 18:48 |
| waynr | and sorry i could not understand both | 18:48 |
| waynr | and be one developer, long i stood | 18:49 |
| waynr | and looked down one as far as I could | 18:49 |
| waynr | to where it bent in the undergrowth | 18:49 |
| waynr | then gave up and worked on something else | 18:49 |
| waynr | (ref: http://www.bartleby.com/119/1.html) | 18:51 |
| clarkb | gertty can really help with that because it is one of the lacking bits of the gerrit ui. It got worse in yhe new change screen too :( | 18:52 |
| clarkb | but people are working to make it better so thats nice | 18:52 |
| waynr | cool | 18:52 |
| *** rfolco has quit IRC | 19:57 | |
| *** baoli has quit IRC | 20:17 | |
| *** cdelatte has quit IRC | 20:24 | |
| *** zxiiro is now known as zxiiro-away | 20:48 | |
| *** baoli has joined #openstack-sprint | 21:48 | |
| *** baoli has quit IRC | 21:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!