| *** lyr has quit IRC | 00:17 | |
| *** yoctozepto has quit IRC | 00:17 | |
| *** zbr has quit IRC | 00:17 | |
| *** zbr has joined #zuul | 00:17 | |
| *** lyr has joined #zuul | 00:17 | |
| *** yoctozepto has joined #zuul | 00:17 | |
| *** msuszko has joined #zuul | 01:25 | |
| *** dry has quit IRC | 01:25 | |
| *** nils has quit IRC | 02:33 | |
| *** jamesmcarthur has quit IRC | 03:12 | |
| *** jamesmcarthur has joined #zuul | 03:14 | |
| *** ricolin has joined #zuul | 03:14 | |
| *** jamesmcarthur has quit IRC | 03:18 | |
| *** jamesmcarthur has joined #zuul | 03:43 | |
| *** ykarel_ has joined #zuul | 03:48 | |
| *** ykarel_ has quit IRC | 03:51 | |
| *** ykarel_ has joined #zuul | 03:51 | |
| *** ykarel_ is now known as ykarel | 03:59 | |
| *** bhavikdbavishi has joined #zuul | 04:04 | |
| *** vishalmanchanda has joined #zuul | 04:43 | |
| *** bhavikdbavishi has quit IRC | 04:46 | |
| *** bhavikdbavishi has joined #zuul | 05:28 | |
| *** jamesmcarthur has quit IRC | 05:50 | |
| *** jamesmcarthur has joined #zuul | 06:03 | |
| *** jamesmcarthur has quit IRC | 06:08 | |
| *** logan- has quit IRC | 06:12 | |
| *** logan- has joined #zuul | 06:18 | |
| *** jamesmcarthur has joined #zuul | 06:21 | |
| *** jamesmcarthur has quit IRC | 06:27 | |
| openstackgerrit | Felix Edel proposed zuul/zuul master: Move "create|delete reference ..." messages to own sub logger https://review.opendev.org/c/zuul/zuul/+/791638 | 06:28 |
|---|---|---|
| *** jamesmcarthur has joined #zuul | 06:41 | |
| *** jamesmcarthur has quit IRC | 06:47 | |
| *** bhavikdbavishi has quit IRC | 06:52 | |
| *** bhavikdbavishi has joined #zuul | 06:52 | |
| *** nils has joined #zuul | 06:56 | |
| *** bhavikdbavishi has quit IRC | 06:57 | |
| *** jamesmcarthur has joined #zuul | 07:01 | |
| *** jamesmcarthur has quit IRC | 07:05 | |
| *** iurygregory has quit IRC | 07:12 | |
| *** hashar has joined #zuul | 07:17 | |
| *** saneax has joined #zuul | 07:19 | |
| *** jamesmcarthur has joined #zuul | 07:20 | |
| *** iurygregory has joined #zuul | 07:21 | |
| *** jamesmcarthur has quit IRC | 07:24 | |
| *** jcapitao has joined #zuul | 07:27 | |
| *** bhavikdbavishi has joined #zuul | 07:28 | |
| *** tosky has joined #zuul | 07:32 | |
| *** bhavikdbavishi has quit IRC | 07:32 | |
| *** jamesmcarthur has joined #zuul | 07:37 | |
| *** jamesmcarthur has quit IRC | 07:44 | |
| *** jamesmcarthur has joined #zuul | 07:57 | |
| *** jpena|off is now known as jpena | 07:58 | |
| *** jamesmcarthur has quit IRC | 08:04 | |
| *** holser has quit IRC | 08:09 | |
| *** ykarel is now known as ykarel|lunch | 08:13 | |
| *** holser has joined #zuul | 08:13 | |
| *** jamesmcarthur has joined #zuul | 08:18 | |
| *** jamesmcarthur has quit IRC | 08:22 | |
| *** jamesmcarthur has joined #zuul | 08:40 | |
| *** sean-k-mooney has quit IRC | 08:47 | |
| *** sean-k-mooney has joined #zuul | 08:53 | |
| *** sean-k-mooney has quit IRC | 09:05 | |
| *** sean-k-mooney has joined #zuul | 09:18 | |
| *** ykarel|lunch is now known as ykarel | 09:27 | |
| *** sean-k-mooney1 has joined #zuul | 09:39 | |
| *** jamesmcarthur has quit IRC | 09:39 | |
| *** sean-k-mooney has quit IRC | 09:40 | |
| *** hashar is now known as hasharAway | 09:58 | |
| *** jpena is now known as jpena|off | 10:58 | |
| *** jpena|off is now known as jpena | 11:00 | |
| *** jcapitao is now known as jcapitao_lunch | 11:04 | |
| *** jpena is now known as jpena|lunch | 11:30 | |
| *** jamesmcarthur has joined #zuul | 11:37 | |
| *** yoctozepto has quit IRC | 11:55 | |
| *** ykarel has quit IRC | 11:55 | |
| *** ykarel has joined #zuul | 12:03 | |
| *** Shrews has joined #zuul | 12:06 | |
| *** hasharAway has quit IRC | 12:09 | |
| *** jcapitao_lunch is now known as jcapitao | 12:09 | |
| *** hashar has joined #zuul | 12:10 | |
| *** sean-k-mooney1 is now known as sean-k-mooney | 12:12 | |
| *** jpena|lunch is now known as jpena | 12:25 | |
| *** yoctozepto has joined #zuul | 12:31 | |
| *** bhavikdbavishi has joined #zuul | 13:14 | |
| *** jamesmcarthur has quit IRC | 13:24 | |
| *** jamesmcarthur has joined #zuul | 13:25 | |
| *** bhavikdbavishi has quit IRC | 13:30 | |
| *** hashar is now known as hasharAway | 14:25 | |
| *** vishalmanchanda has quit IRC | 14:54 | |
| *** adpopescu has joined #zuul | 15:05 | |
| *** jcapitao has quit IRC | 15:10 | |
| *** jcapitao has joined #zuul | 15:10 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Be explicit about yaml loaders/dumpers https://review.opendev.org/c/zuul/zuul/+/791765 | 15:25 |
| corvus | clarkb: ^ there's a followup to your change which eliminates the metaclass usage | 15:25 |
| corvus | clarkb: i'll propose the unrevert on top of that, and i think we can be fairly certain that production will actually behave like tests next time | 15:26 |
| *** jcapitao has quit IRC | 15:29 | |
| *** jcapitao has joined #zuul | 15:30 | |
| corvus | mordred: ^ | 15:30 |
| clarkb | corvus: Ithink that change may have the issue my earlier patchset had which is that we set SafeLoader conditionally based on import stuff at the top of the file so we don't want to use yaml.SafeLoader elsewhere and instead should use SafeLoader. Otherwise, I like the way that change looks and everthing else uses the encrypted loader | 15:32 |
| clarkb | the good news is I'm 99% sure your change will work as is without that fixup and that is because you are explicit :) | 15:34 |
| corvus | clarkb: oh yep; it does work, but i'll fix it up to use 'SafeLoader' in a minute | 15:35 |
| mordred | corvus: ++ ... change looks good to me! (pending the SafeLoader thing) | 15:37 |
| *** ykarel has quit IRC | 15:39 | |
| corvus | clarkb, mordred: i just applied the unrevert without updating to use the new methods and ran a test locally and it correctly failed (just as we saw in production)! | 15:41 |
| mordred | corvus: yay! | 15:41 |
| clarkb | excellent | 15:42 |
| fungi | that's great news | 15:45 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Be explicit about yaml loaders/dumpers https://review.opendev.org/c/zuul/zuul/+/791765 | 15:45 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Revert "Omnibus executor secret decrypt revert" https://review.opendev.org/c/zuul/zuul/+/791775 | 15:45 |
| corvus | okay that should be the whole shebang | 15:46 |
| mordred | corvus: on the revert - there's a EncryptedDumper.add_representer(types.MappingProxyType, | 15:48 |
| mordred | yaml.representer.SafeRepresenter.represent_dict) | 15:48 |
| mordred | that pulls from yaml directly rather than the cyaml/yaml - is that ok? | 15:48 |
| corvus | mordred: i think so, lemme double check | 15:49 |
| mordred | otherwise - stack looks great | 15:49 |
| corvus | mordred: yeah, a quick browse through https://github.com/yaml/pyyaml/tree/master/yaml and i don't see a definition there, and https://github.com/yaml/pyyaml/blob/master/lib3/yaml/cyaml.py is just re-importing the python versions for that class -- but why don't i update it anyway just to be consistent and maybe future-proof? | 15:53 |
| corvus | oh actually | 15:53 |
| corvus | it looks like there's a bunch of other yaml.* stuff in that file | 15:54 |
| corvus | so maybe it's more consistent to leave as-is | 15:54 |
| mordred | makes sense - those are object model things rather than actually serialization related - so don't really need to be different for cyaml | 15:54 |
| corvus | oh, now on a third look i see that there's some inconsistency there related to the move we just made | 15:56 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Revert "Omnibus executor secret decrypt revert" https://review.opendev.org/c/zuul/zuul/+/791775 | 16:01 |
| clarkb | corvus: once ^ lands is it possible to restart a single executor to check it is working properly or do we also need to restart the scheduler? | 16:11 |
| clarkb | wondering if we can get away with a smaller scale asnity check | 16:11 |
| openstackgerrit | Zane Bitter proposed zuul/zuul master: docs: Fix spelling mistake https://review.opendev.org/c/zuul/zuul/+/791800 | 16:14 |
| *** Shrews has quit IRC | 16:21 | |
| corvus | clarkb: gotta be both | 16:31 |
| *** jcapitao has quit IRC | 16:42 | |
| *** jamesmcarthur has quit IRC | 16:51 | |
| melwitt | clarkb: I dunno if you saw but I updated the build page button patch with a way to only show it when the top of page isn't visible https://review.opendev.org/c/zuul/zuul/+/790319 | 16:52 |
| *** jamesmcarthur has joined #zuul | 17:00 | |
| *** jpena is now known as jpena|off | 17:01 | |
| *** jamesmcarthur_ has joined #zuul | 17:04 | |
| *** jamesmcarthur has quit IRC | 17:08 | |
| *** Shrews has joined #zuul | 17:15 | |
| *** hamalq has joined #zuul | 17:25 | |
| openstackgerrit | Merged zuul/zuul master: docs: Fix spelling mistake https://review.opendev.org/c/zuul/zuul/+/791800 | 17:28 |
| clarkb | melwitt: I hadn't seen that, nice | 17:28 |
| *** simondodsley has joined #zuul | 17:55 | |
| corvus | mordred, clarkb: interestingly, i think the s/yaml.// change may have altered the output slightly -- the original produced quoted strings and the update unquoted. | 17:56 |
| corvus | it's no big deal -- but that caused the tests to fail since it didn't match | 17:56 |
| clarkb | a behavior difference between libyaml and pyyaml? both valid yaml though | 17:57 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Be explicit about yaml loaders/dumpers https://review.opendev.org/c/zuul/zuul/+/791765 | 17:59 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Revert "Omnibus executor secret decrypt revert" https://review.opendev.org/c/zuul/zuul/+/791775 | 17:59 |
| simondodsley | Newbie question... Using Zuul for Devstack, how do I tell git to clone the branch required? specifically with the Manila project this always seems to get master even when all other OpenStack projects pick up the correct stable branch. I need to pass something like $MANILA_BRANCH to it, but how do I populate that? Is there an ansible variable with that informatino already pre-populated? | 17:59 |
| corvus | clarkb: seems that way | 17:59 |
| corvus | clarkb, mordred: if you want to go ahead and re-review those two -- that was the only test failure; i think we can go ahead and approve them | 17:59 |
| clarkb | corvus: I'll put that on the list after the swap cleanups | 18:00 |
| clarkb | simondodsley: that may be a better question for the devsatck maintainers in #openstack-qa as they maintain the jobs and their interfaces too. But generally zuul will checkout the branch for the change that was proposed | 18:00 |
| clarkb | simondodsley: this means if you are in a change context it is automatic for you. If outside of a change context (periodic or pulling in something special) then there are override vars | 18:01 |
| simondodsley | clarkb: i see that is the case for most projects, but for devstack manila is installed as a plugin. I'll go ask elsewhere as suggested | 18:01 |
| *** hasharAway is now known as hashar | 18:12 | |
| clarkb | corvus: I know this is an unrevert but why do the parameters change in https://review.opendev.org/c/zuul/zuul/+/791775/3/zuul/executor/common.py ? do we expect that to break any existing jobs? | 18:29 |
| clarkb | corvus: looks like we expect jobs to use the agent instead? | 18:30 |
| clarkb | other than that the unrevert looks about how I expceted it to. I didn't approve anything as the bottom two changes need review still | 18:37 |
| *** jamesmcarthur_ has quit IRC | 18:44 | |
| mordred | corvus: I bet the quote/non-quote change in the test shows that there was a codepath where we weren't using libyaml before that patch | 18:46 |
| mordred | corvus: also - the stack is now 2x+2 - didn't +A in case that wants to be managed | 18:47 |
| corvus | clarkb: params change because the meaning is changing. that's how the project ssh keys were passed to the executor. in both cases (old/new) ssh_keys is a list of dicts. the old dict was the name of the ssh key and the private key itself. the new dict is the connection and project name of the ssh key -- enough information for the executor to find the private key and decrypt it. it'll be responsible for | 19:05 |
| corvus | coming up with a name. | 19:05 |
| corvus | clarkb: here's where it generates the key name now: https://review.opendev.org/c/zuul/zuul/+/791775/3/zuul/executor/server.py#954 | 19:06 |
| corvus | clarkb: in all cases, the executor is still adding the key to the agent. the difference is where it gets the private key (old=decrypted via gearman, new=encrypted via zk) | 19:07 |
| clarkb | oh got it its the lookup info | 19:07 |
| corvus | yep | 19:07 |
| clarkb | but then on the job side nothing has chagned | 19:07 |
| corvus | exactly :) | 19:07 |
| corvus | mordred: ack, and thanks. i've +wd it now | 19:08 |
| *** jamesmcarthur has joined #zuul | 19:10 | |
| clarkb | zuulians, is the `data['tag'] = 'autogenerated:zuul:%s' % (item.pipeline.name)` attribute of zuul http reports the thing that allows those comments to be filtered by the gerrit "Only Comments" filter? | 19:32 |
| corvus | clarkb: i believe so; specifically the 'autogenerated' part | 19:33 |
| clarkb | ya I think I just foudn a wikimedia issue that confirms | 19:33 |
| corvus | speaking of -- clarkb, ianw: do you think we could have the zuul plugin filter out comments automatically based on tags? | 19:34 |
| corvus | i don't think we should auto-remove all zuul comments | 19:34 |
| corvus | but, we could tag comments which contain nothing other than stuff that shows up in the summary tab, and filter *those* out | 19:35 |
| corvus | (we do want to keep showing error/warning messages, etc) | 19:35 |
| corvus | i think we talked about this briefly before, but i suspect folks may know more about plugin implementations now | 19:35 |
| fungi | errors/warnings may be worth continuing to expose. we've had plenty of cases in opendev in times past where people didn't spot errors because they didn't look at zuul's review comment text due to us hiding it with javascript toggles | 19:36 |
| fungi | though if gerrit provides a means of better raising those error messages, maybe that's even better than comments | 19:36 |
| corvus | yeah. filtering out the noise will make it easier for folks to spot actual errors | 19:36 |
| clarkb | ya I'm asking bceause the cinder team is wondering how they can get more CI systems to be like zuul and have only comments filter and the status plugin results show up for third party ci | 19:36 |
| clarkb | they put it on the agenda for tomorrow so we can discuss if they think that would help too. | 19:37 |
| corvus | they could run zuul :) | 19:37 |
| clarkb | corvus: I'm literally writing that as a response to their question in the wiki :) | 19:37 |
| fungi | in many cases i think they *do* run zuul, just too old a zuul to have that feature | 19:37 |
| fungi | (many still 2.x) | 19:40 |
| *** hashar has quit IRC | 19:50 | |
| openstackgerrit | Merged zuul/zuul master: Move encrypted/pkcs1-oaep yaml object handling into yamlutil https://review.opendev.org/c/zuul/zuul/+/791514 | 21:16 |
| openstackgerrit | Merged zuul/zuul master: Be explicit about yaml loaders/dumpers https://review.opendev.org/c/zuul/zuul/+/791765 | 21:19 |
| openstackgerrit | Merged zuul/zuul master: Revert "Omnibus executor secret decrypt revert" https://review.opendev.org/c/zuul/zuul/+/791775 | 21:19 |
| -openstackstatus- NOTICE: The Zuul service at zuul.opendev.org will be offline for a few minutes (starting now) in order for us to make some needed filesystem changes; if the outage lasts longer than anticipated we'll issue further notices | 21:34 | |
| corvus | zuul-maint: opendev is now running with that change ^ so master is looking good so far | 22:10 |
| *** iurygregory has quit IRC | 22:10 | |
| *** nils has quit IRC | 22:11 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Execute builds via ZooKeeper https://review.opendev.org/c/zuul/zuul/+/788988 | 22:17 |
| *** iurygregory has joined #zuul | 22:19 | |
| *** jamesmcarthur has quit IRC | 22:20 | |
| *** jamesmcarthur has joined #zuul | 22:21 | |
| *** jamesmcarthur_ has joined #zuul | 22:22 | |
| mordred | corvus: yay for sucess! | 22:23 |
| corvus | mordred: there's no success like success | 22:25 |
| *** jamesmcarthur has quit IRC | 22:26 | |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Improve component registry https://review.opendev.org/c/zuul/zuul/+/787684 | 22:27 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Calculate statsd metrics via ComponentRegistry https://review.opendev.org/c/zuul/zuul/+/787685 | 22:27 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Add ExecutorApi https://review.opendev.org/c/zuul/zuul/+/770902 | 22:27 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Change zone handling in ExecutorApi https://review.opendev.org/c/zuul/zuul/+/787833 | 22:27 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Clean up Executor API build request locking and add tests https://review.opendev.org/c/zuul/zuul/+/788624 | 22:27 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Execute builds via ZooKeeper https://review.opendev.org/c/zuul/zuul/+/788988 | 22:27 |
| mordred | corvus: I thought failure was a great way to succeed, no? | 22:28 |
| corvus | mordred: maybe redefining success is the best way? | 22:30 |
| mordred | corvus: oh - yeah. redefining success is absolutely the best way to success | 22:30 |
| ianw | corvus: hrm i still don't know *that* much about plugins :) the "only comments" slider though does remove zuul comments, right? | 22:34 |
| corvus | ianw: yep. i think the ideal would be not to have 'boring' comments show up in the first place. we don't have a tag to indicate that, but if it's possible, then we could add it. | 22:35 |
| corvus | not all zuul comments are boring; but all the comments that contain only data that are extracted into the zuul summary tab are boring | 22:35 |
| ianw | it looks like they removed the "only comments" slider anyway -> https://gerrit-review.googlesource.com/c/gerrit/+/274277 | 22:42 |
| corvus | all the more reason to see if we can do it | 22:45 |
| corvus | clarkb: ^ fyi | 22:45 |
| corvus | is that not in the release opendev is running? | 22:45 |
| clarkb | we have the only comments toggle in our gerrit | 22:46 |
| clarkb | maybe they undid the undoing? | 22:46 |
| corvus | clarkb: the periodic job error you reported in #opendev is the same error that was recently posted to the list regarding 'does not have the default branch master' | 22:51 |
| corvus | clarkb: that queue item is the result of a re-enqueue from fungi, right? | 22:51 |
| clarkb | yes | 22:51 |
| clarkb | should we dequeue it? | 22:51 |
| corvus | nah, it's harmless | 22:51 |
| clarkb | roger | 22:51 |
| corvus | it will just continue to fail once that last job gets a node | 22:52 |
| *** jamesmcarthur_ has quit IRC | 22:52 | |
| corvus | but when zuul really enqueues it with the timer trigger it should work | 22:52 |
| clarkb | does that mean periodic jobs can no longer be enqueued this way? maybe they couldn't before too and I didn't notice | 22:53 |
| corvus | so relevant to #zuul -- confirmation of bug report on mailing list (not that we needed it, i believed the report :) and relevant to #opendev, since it's peridoc jobs and they're not critical, i would ignore it and just let it get fixed tomorrow. if necessary, it can be enqueued with an explicit sha, but that's a lot of work. | 22:53 |
| clarkb | wfm | 22:54 |
| corvus | clarkb: right now that's the case, but it's a regression; we need to fix zuul. (it's not an intended behavior change) | 22:54 |
| corvus | clarkb: i *think* but am not certain, that it's related to the global repo state | 22:54 |
| corvus | at least, inasmuch as i think that's where it snuck in. | 22:55 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Improve component registry https://review.opendev.org/c/zuul/zuul/+/787684 | 22:55 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Calculate statsd metrics via ComponentRegistry https://review.opendev.org/c/zuul/zuul/+/787685 | 22:55 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Add ExecutorApi https://review.opendev.org/c/zuul/zuul/+/770902 | 22:55 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Change zone handling in ExecutorApi https://review.opendev.org/c/zuul/zuul/+/787833 | 22:55 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Clean up Executor API build request locking and add tests https://review.opendev.org/c/zuul/zuul/+/788624 | 22:55 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Execute builds via ZooKeeper https://review.opendev.org/c/zuul/zuul/+/788988 | 22:55 |
| clarkb | corvus: ianw re https://gerrit-review.googlesource.com/c/gerrit/+/274277 that change seems to be part of them changing it from an experimental feature to an actual feature | 22:57 |
| clarkb | I don't think they are removing it | 22:57 |
| clarkb | it was just code cleanup after they promoted that feature from experimental | 22:57 |
| corvus | if we wanted to experiment, we could push up a dnm change with a zuul config error and get it to leave a message :) | 22:59 |
| ianw | clarkb: yeah, i can't find that element defined in the master tree. unless somehow it moved into a plugin | 23:01 |
| clarkb | ianw: gr-messages-list or gr-messages-list-experimental? | 23:01 |
| clarkb | -experimental does appear to be removed | 23:01 |
| ianw | my path so far has been to "inspect element" on our gerrit, which led me to id=automatedMessageToggle, which led me to that change. but i can't find references to automatedMessageToggle any more | 23:02 |
| ianw | https://gerrit-review.googlesource.com doens't have the toggle either atm | 23:03 |
| clarkb | gr-messages-list shows up in the repo at least | 23:04 |
| clarkb | btu ya may not do what I expect it to anymore? | 23:04 |
| clarkb | or I'm misinterpreting the log | 23:04 |
| corvus | i checked a 3.3 gerrit and i don't see it (i also don't see 'comment threads' tab) | 23:14 |
| ianw | gr-message.ts appears to still have a property "hideAutomated = false;" but i can see nothing that ever flips it to be true | 23:17 |
| fungi | maybe they expect that to be handled by a pg plugin now? | 23:18 |
| openstackgerrit | Merged zuul/zuul master: Calculate build start and end time on executor server https://review.opendev.org/c/zuul/zuul/+/788915 | 23:21 |
| ianw | ok, it changed to "Show all entries" | 23:24 |
| ianw | https://gerrit-review.googlesource.com/c/gerrit/+/264181 | 23:25 |
| ianw | and "Show all entries" does appear on that page | 23:25 |
| clarkb | ianw: maybe it only pops up if it has hidden entries? | 23:26 |
| clarkb | the Show all entries stuff still seems to be in gerrit master .ts file | 23:27 |
| ianw | _isVisibleShowAllActivityToggle(messages: CombinedMessage[] = []) { | 23:27 |
| ianw | return messages.some(m => !m.isImportant); | 23:27 |
| ianw | so it's only visible if some messages are not important if i'm reading that right | 23:28 |
| clarkb | heh so now we need to figure out how messages are annotated important | 23:28 |
| clarkb | and then have zuul mark its as not important if that is different than what zuul alerady does | 23:28 |
| ianw | if (!message.tag) return true; | 23:29 |
| ianw | function computeIsImportant( | 23:29 |
| clarkb | cool so that means we should continue to have the same functionality just under a different name | 23:29 |
| clarkb | also looks like it trims messages for older patchsets by default | 23:30 |
| ianw | so ... if it does not have a message.tag, then it's considered important; and then yeah, has some logic for pruning old revisions | 23:30 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Clean up Executor API build request locking and add tests https://review.opendev.org/c/zuul/zuul/+/788624 | 23:30 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Execute builds via ZooKeeper https://review.opendev.org/c/zuul/zuul/+/788988 | 23:30 |
| openstackgerrit | James E. Blair proposed zuul/zuul master: Switch to string constants in BuildRequest https://review.opendev.org/c/zuul/zuul/+/791849 | 23:30 |
| clarkb | https://gerrit-review.googlesource.com/c/plugins/healthcheck/+/304528 has appropriate state to enable the toggle | 23:31 |
| corvus | interesting... does it also need to have a lot of entries? | 23:31 |
| clarkb | corvus: didn't look like it filtered by count I think something else may do that trimming | 23:32 |
| corvus | because that did not show up on my 3.3 server | 23:32 |
| ianw | it needs to have some "not important" messages to show the toggle | 23:32 |
| ianw | https://gerrit.googlesource.com/gerrit/+/refs/heads/master/polygerrit-ui/app/elements/change/gr-messages-list/gr-messages-list.ts#187 is "isImportant" | 23:34 |
| ianw | https://gerrit.googlesource.com/gerrit/+/refs/heads/master/polygerrit-ui/app/elements/change/gr-messages-list/gr-messages-list.ts#426 then controls the toggle showing up | 23:34 |
| corvus | i read clarkb's message as saying that zuul autogenerated tagged messages were 'not important' | 23:34 |
| corvus | are we unsure whether that's true? | 23:35 |
| ianw | a tagged message is not important, that's #187 link above | 23:35 |
| ianw | so if zuul messages wanted to show up as important, they should not be tagged, i guess | 23:36 |
| ianw | if particular messages i mean, messages that are ... important :) | 23:36 |
| corvus | the only zuul messag getting filtered out of that is the one from the earlier patchset | 23:36 |
| corvus | but then, it's not leaving a textual message, so i don't know if there's actually a tag attached | 23:37 |
| corvus | on my gerrit 3.3 server, zuul does leave messages (just like on opendev), and i don't see the slider | 23:37 |
| corvus | so something still isn't adding up for me | 23:37 |
| corvus | maybe master has different behavior than 3.3 | 23:38 |
| ianw | * Autogenerated messages are unimportant, if there is a message with the same | 23:39 |
| ianw | * tag and a higher revision number. | 23:39 |
| corvus | ah, that's helpful; lemme see if i can find that condition in my 3.3 | 23:40 |
| corvus | ianw, clarkb: ah yep, 3.3 behaves like master in that case | 23:42 |
| *** tosky has quit IRC | 23:43 | |
| corvus | so you'll see the most recent zuul spam per pipeline by default. that's not bad. | 23:43 |
| ianw | so the latest Zuul comment (even when tagged) should be showing up | 23:43 |
| corvus | per-pipeline (since the tag includes the pipeline) but yes | 23:43 |
| ianw | yep | 23:43 |
| corvus | and i think that's exactly what we'd want too :) | 23:43 |
| corvus | don't want 'experimental' burying 'check' | 23:44 |
| corvus | (if we wanted it to show up at all) | 23:44 |
| corvus | so there's still the left-field idea of doing more filtering in the zuul plugin (because we could hide comments which only have info in the zuul tab). i'm still curious if that's possible :) | 23:45 |
| ianw | currently the functions only fire when the tab is clicked on | 23:50 |
| corvus | ok, so not an easy addition to the current scheme | 23:51 |
| ianw | we've have to write something that listens for different end-points -> https://gerrit-review.googlesource.com/Documentation/pg-plugin-endpoints.html#_plugin_endpoints | 23:54 |
| ianw | i'm not actualy sure if you can modify the comments list being displayed via a plugin | 23:56 |
| *** hamalq has quit IRC | 23:56 | |
| *** hamalq has joined #zuul | 23:57 | |
| ianw | for those hooks, you'll get passed an instance of ChangeInfo which has all your comments, etc. which you can use to draw your own things | 23:57 |
| ianw | but I don't think a plugin can really affect what gr-messages-list/gr-messages-list.ts is drawing | 23:57 |
| ianw | modulo the plugin doing something like monkey-patching internal functions of gr-message-list.ts like "isImportant" etc. ... which i bet is totally possible and also horrendously ugly :) | 23:59 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!