| *** tetsuro has joined #openstack-tc | 00:30 | |
| *** tetsuro_ has joined #openstack-tc | 00:40 | |
| *** tetsuro has quit IRC | 00:43 | |
| mnaser | clarkb, fungi: given your intimate knowledge of our infra, do you know how much 'extra load' we would add if projects actually tested for 1 or 2 more distros (say debian and centos) -- and probably one job only. is the impact big enough or it won't really be noticed? | 00:52 |
|---|---|---|
| clarkb | mnaser: its been a while since I ran the generation of those stats but they are in graphite now so anyone should be able to look at them I think | 01:02 |
| clarkb | in general ot seems that the job cost itself is a major factor which is why tripleo has been a major consumer | 01:02 |
| clarkb | long running jobs with multiple nodes consume a lot relatove to other jobs | 01:03 |
| clarkb | all that to say it really depends on the type of jobs you want to run | 01:03 |
| mnaser | clarkb: right, there seems to be a lot of discussion with the "Remove centos 8 from the list" but in reality we don't really test any of the othrs so they're not more deserving to be there than the others | 01:03 |
| mnaser | so wondering how much extra work we might end up needing if we were to implement something so all projects run at least a _single_ devstack based job on all the supported platforms | 01:04 |
| clarkb | a single node devstack probably wouldnt be too bad. You'd need people to keep devstack running on those platforms though | 01:05 |
| clarkb | also note fungi pointed out earlier the docs dont actually say w e must test thoseplatofrms | 01:06 |
| clarkb | just that those platforms are common and projects should test them if they can | 01:06 |
| mnaser | yeah, i agree. it just seems like it's a big point of contention right now amongst people that consider those platforms "untestable" | 01:06 |
| mnaser | i'm almost tempted to rip out that entire document in that case because we just ship a bunch of python code. the fact we test it on ubuntu validates that it works, but not that specific operating systems are supported | 01:07 |
| gmann | also stability of job will be more important, we need to check it would not slow down the development | 01:07 |
| clarkb | really itslong running multinode jobs that have the biggest impact | 01:08 |
| clarkb | single node devstack is probably fine but we can run actual numbers and quantify it (just not this moment as it is dinner time) | 01:08 |
| mnaser | gmann: yeah i imagine from a qa perspective we probably have the whole ubuntu platform nailed down | 01:22 |
| mnaser | as someone whos done a lot of deployment across different os, generally, platforms are pretty stable | 01:22 |
| gmann | mnaser: 1 idea is to add periodic jobs on each project that can test each distro daily basis and would not have much impact on current gate situation. | 01:25 |
| gmann | like we do for oslo master testing in "tempest-full-oslo-master" | 01:25 |
| mnaser | gmann: true but in my experience no one ends up paying attention to periodic jobs :X | 01:26 |
| gmann | that depends on team to team but for majority i agree those are not actively monitored . | 01:27 |
| mnaser | yeah i generally feel like those go usually unnoticed | 01:29 |
| gmann | so how much testing we want to cover ? per services only or per repo (mean all the code changes including for client, lib etc ) and functional tests only or integration tests too | 01:29 |
| gmann | integration tests might question on stability of job (like ubuntu based are also not 100% stable) as functional testing is much stable. | 01:31 |
| mnaser | gmann: i think you'd probably be the best domain expert in that (or maybe we should stop promising "supported operating systems" because the list doesn't really mean much) | 01:34 |
| gmann | my initial plan/arguments was always to tests all listed distro but it was our doc which was/is confusing for me on tested vs supported vs list to decide the testing runtime. | 01:37 |
| gmann | mnaser: let me try out some devstack jobs with some other services than what we have on devstack and get some stat. and then we can decide if to test those distro or stop promising their support. | 01:38 |
| mnaser | perhaps. i mean, if we drop the OS, we should drop all operating systems as "tested" and make no promises about the OS that we test on | 01:39 |
| mnaser | it's a little unfair for other distros just because ubuntu was the day 1 resident that we test on (i.e. i'd personally have preferred it to be debian but it is what it is!) | 01:40 |
| gmann | that's fair point. | 01:40 |
| *** tetsuro has joined #openstack-tc | 01:44 | |
| *** tetsuro_ has quit IRC | 01:47 | |
| *** tetsuro_ has joined #openstack-tc | 02:12 | |
| *** tetsuro has quit IRC | 02:15 | |
| fungi | back to my earlier point, we need some way of deciding when we can drop support for specific versions of some dependencies | 02:24 |
| fungi | how we've been deciding that is by looking at what's shipped in lts server distros on which our users commonly deploy our software | 02:25 |
| fungi | keeping a list of the distros we use in that calculation is the main benefit i see to that bit of the document | 02:25 |
| fungi | otherwise how do we determine, for example, when it's safe to start releasing code which can no longer run under python 3.6? | 02:26 |
| fungi | that's what the list of distros is there for, and when that section was added we made a careful attempt to avoid saying they were "supported" or that all our software was "tested" on them | 02:27 |
| fungi | it was taken as an aspirational goal to suggest that projects should run functional tests on those platforms, that was acknowledged in the review comments as well, and as was observed at the time could form the basis of some cycle goals around increased testing | 02:30 |
| fungi | but given that most projects haven't added their own functional tests on additional platforms, four years later, it seems like we're better off removing that suggestion | 02:31 |
| *** tetsuro has joined #openstack-tc | 03:06 | |
| *** tetsuro_ has quit IRC | 03:09 | |
| *** tetsuro_ has joined #openstack-tc | 04:06 | |
| *** tetsuro has quit IRC | 04:09 | |
| *** evrardjp has quit IRC | 04:36 | |
| *** evrardjp has joined #openstack-tc | 04:36 | |
| *** diablo_rojo has quit IRC | 04:44 | |
| *** dmellado has quit IRC | 04:45 | |
| *** dklyle has quit IRC | 05:01 | |
| *** tetsuro has joined #openstack-tc | 05:43 | |
| *** tetsuro_ has quit IRC | 05:47 | |
| *** njohnston has quit IRC | 06:05 | |
| *** tetsuro_ has joined #openstack-tc | 06:23 | |
| *** belmoreira has joined #openstack-tc | 06:24 | |
| *** tetsuro has quit IRC | 06:26 | |
| *** tetsuro has joined #openstack-tc | 06:29 | |
| *** tetsuro_ has quit IRC | 06:32 | |
| *** lpetrut has joined #openstack-tc | 06:42 | |
| *** slaweq has joined #openstack-tc | 06:51 | |
| *** witek_ has joined #openstack-tc | 06:59 | |
| *** tetsuro_ has joined #openstack-tc | 07:24 | |
| *** tetsuro has quit IRC | 07:27 | |
| *** rpittau|afk is now known as rpittau | 07:29 | |
| *** tosky has joined #openstack-tc | 07:35 | |
| *** dmellado has joined #openstack-tc | 07:37 | |
| *** e0ne has joined #openstack-tc | 08:05 | |
| *** tetsuro_ has quit IRC | 08:47 | |
| *** tetsuro has joined #openstack-tc | 09:07 | |
| *** tetsuro has quit IRC | 09:23 | |
| openstackgerrit | Graham Hayes proposed openstack/governance master: Require PTL signoff for project-update changes https://review.opendev.org/727704 | 10:41 |
| *** njohnston has joined #openstack-tc | 11:01 | |
| *** rpittau is now known as rpittau|bbl | 11:04 | |
| *** rpittau|bbl is now known as rpittau | 12:19 | |
| *** tkajinam has quit IRC | 12:24 | |
| *** lpetrut has quit IRC | 12:46 | |
| *** ianychoi has quit IRC | 12:55 | |
| jungleboyj | o/ | 13:06 |
| njohnston | o/ | 13:09 |
| belmoreira | o/ | 13:34 |
| mnaser | excellent patch mugsie | 13:38 |
| gmann | o/ | 13:50 |
| *** lpetrut has joined #openstack-tc | 14:29 | |
| knikolla | o/ | 14:31 |
| *** dklyle has joined #openstack-tc | 14:42 | |
| gmann | njohnston: on V cycle goal selection, its time to do now. I will say we(tc) decide on IRC instead of gerrit vote what yo say? or any better and fast idea | 14:45 |
| cloudnull | o/ | 14:56 |
| evrardjp | gmann: WFM | 14:58 |
| gmann | evrardjp: cool | 14:59 |
| gmann | tc-members: let's discuss V cycle goal selection and finalise it tomorrow office hour | 15:00 |
| gmann | or in meeting, mnaser when is our meeting next week? | 15:01 |
| mnaser | gmann: i say lets do it for office hours, i will need to figure out meeting logistics. the whole end of cycle thing + getting back to chair duties is slowly getting us started | 15:02 |
| gmann | mnaser: ok | 15:02 |
| mnaser | gmann: i feel like this needs a non-async discussion too | 15:03 |
| mnaser | it'll stall out for too long otherwise | 15:03 |
| mnaser | tc-members: who is able to attend tomorrows office hours for ^ ? can we get quorum? | 15:03 |
| knikolla | i will attend. | 15:03 |
| gmann | mnaser: yeah, as ussuri is released officially we need to final it now. | 15:03 |
| gmann | i will be there. | 15:04 |
| *** lpetrut has quit IRC | 15:08 | |
| * njohnston will be there | 15:08 | |
| *** belmoreira has quit IRC | 15:24 | |
| jungleboyj | I will be in office hours tomorrow. Have another call at the same time but should still be relatively dialed in. :-) | 15:24 |
| *** witek_ has quit IRC | 16:15 | |
| *** rpittau is now known as rpittau|afk | 16:21 | |
| *** camelCaser has quit IRC | 16:22 | |
| *** ccamel has joined #openstack-tc | 16:22 | |
| *** witek_ has joined #openstack-tc | 16:26 | |
| *** evrardjp has quit IRC | 16:36 | |
| *** evrardjp has joined #openstack-tc | 16:36 | |
| diablo_rojo_phon | I can be there if you need me for quorum. | 17:11 |
| *** ijolliffe has quit IRC | 17:36 | |
| *** e0ne has quit IRC | 17:55 | |
| *** belmoreira has joined #openstack-tc | 19:05 | |
| *** dklyle has quit IRC | 19:38 | |
| *** dklyle has joined #openstack-tc | 19:46 | |
| *** belmoreira has quit IRC | 20:07 | |
| *** slaweq has quit IRC | 21:57 | |
| *** slaweq has joined #openstack-tc | 22:08 | |
| *** slaweq has quit IRC | 22:13 | |
| *** slaweq has joined #openstack-tc | 22:23 | |
| *** slaweq has quit IRC | 22:28 | |
| *** tkajinam has joined #openstack-tc | 22:46 | |
| *** tosky has quit IRC | 23:06 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!