| *** diablo_rojo has quit IRC | 00:05 | |
| *** timburke has quit IRC | 01:41 | |
| *** timburke_ has joined #openstack-tc | 01:41 | |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #openstack-tc | 05:33 | |
| *** lpetrut has joined #openstack-tc | 05:50 | |
| *** lpetrut has quit IRC | 05:54 | |
| openstackgerrit | Pranali Deore proposed openstack/governance master: Add glance-tempest-plugin to Glance https://review.opendev.org/c/openstack/governance/+/767666 | 06:22 |
|---|---|---|
| *** lpetrut has joined #openstack-tc | 06:28 | |
| *** lpetrut has quit IRC | 06:30 | |
| *** slaweq has joined #openstack-tc | 07:42 | |
| *** dklyle has quit IRC | 07:57 | |
| *** ralonsoh has joined #openstack-tc | 08:02 | |
| *** andrewbonney has joined #openstack-tc | 08:15 | |
| *** rpittau|afk is now known as rpittau | 08:30 | |
| *** e0ne has joined #openstack-tc | 08:34 | |
| *** e0ne has quit IRC | 08:41 | |
| *** slaweq has quit IRC | 09:16 | |
| *** slaweq has joined #openstack-tc | 09:18 | |
| *** slaweq has quit IRC | 09:20 | |
| *** slaweq has joined #openstack-tc | 09:24 | |
| ttx | fungi: all names had some level of issues, but Xena was deemed "safe enough". | 10:12 |
| openstackgerrit | Pranali Deore proposed openstack/governance master: Add glance-tempest-plugin to Glance https://review.opendev.org/c/openstack/governance/+/767666 | 10:42 |
| *** slaweq has quit IRC | 10:51 | |
| *** slaweq has joined #openstack-tc | 10:55 | |
| *** e0ne has joined #openstack-tc | 12:01 | |
| *** tosky has joined #openstack-tc | 12:58 | |
| ttx | mnaser: with AJaeger no longer helping as much on OpenStack project-config reviews, project-config-core is down to you alone. I think it's time to put out a call for new members into support the "openstack consumer" side of our infrastructure | 13:15 |
| ttx | With the remaining infra-root focusing on the opendev side (providing the services to all projects), we need new blood on the openstack side | 13:16 |
| zbr|rover | :( | 13:27 |
| mnaser | ttx: ah, I think I am also not that great at reviewing. More of an “review on demand” sort of thing | 13:29 |
| zbr|rover | mnaser: as I said in the past, i am interested to help in the infra area. | 13:35 |
| mnaser | zbr|rover: thank you, i'm still unsure as to who manages project-config-core now | 13:40 |
| mnaser | i dont think we got much opendev-y governance done so is it up to me to add new members? is it up to tact chair? is it up to opendev who mostly occupy project-config-core? | 13:40 |
| * mnaser shrug | 13:40 | |
| *** e0ne has quit IRC | 14:09 | |
| *** e0ne has joined #openstack-tc | 14:09 | |
| ttx | mnaser: That's what I meant by "openstack needs to start owning that side" | 14:11 |
| ttx | It was a single group, but now that the opendev core services are used by multiple projects, operations of the services needs to be separated from configuration/usage of the services | 14:12 |
| ttx | The former is opendev/centralized. The latter is project-specific | 14:12 |
| ttx | The latter involves owning the OpenStack gerrit projects, OpenStack Zuul jobs, and debugging when "CI fails" | 14:14 |
| ttx | zbr|rover: which infra area makes most sense for you to be involved in? Core opendev operations, or OpenStack's consumption of those? | 14:15 |
| ttx | I feel like today if we/openstack own more of the latter, that will free up time for infra-root to focus on the former | 14:16 |
| ttx | Andreas's crazy review activity made it possible to ignore the issue until today, but now it's urgent | 14:17 |
| ttx | although project-config is likely still a shared property | 14:19 |
| zbr|rover | i was referring to opendev operations, but my impression is that decoupling may not be complete ever | 14:19 |
| zbr|rover | config changes do impact operations and can bring services down, sometimes changes that look genuine can do this. | 14:20 |
| ttx | sure it still requires lots of communications. But at least the core group no longer needs to know by heart all the job structure for all the projects | 14:22 |
| ttx | if they are "owned" by some other group | 14:22 |
| ttx | Also I'm pretty sure that if a project-level config change can bring a service down, that's bug that should ultimately be fixed | 14:23 |
| mnaser | i think the focus of what ttx is talking about is specifically things like openstack/project-config | 14:33 |
| mnaser | like adding a project to zuul/gerrit, etc | 14:33 |
| ttx | yes that and generally knowing the openstack zuul jobs structure / templates /etc. to be able to serve as first line of debugging and help | 14:35 |
| *** iurygregory has joined #openstack-tc | 14:38 | |
| zbr|rover | mnaser: i guess that giving more access to project-config is more accessible, but at the same time i think that anyone that gets core permissions to project-config should be fully trained to debug issues with the runnings services (read only access), so they can debug if needed. | 14:40 |
| mnaser | ttx, zbr|rover: i guess this is worth adding as a discussion item for next weekly tc discussion | 15:00 |
| fungi | if it helps inform the situation, andreas did not have (nor want) root access to any running services | 15:00 |
| mnaser | fungi: right, yeah, i think we're mixing infra-root and infra-core a tad bit here | 15:01 |
| fungi | but managed to review massive amounts of project-config change activity | 15:01 |
| fungi | i was responding to zbr|rover's suggestion that anyone with config core reviewer access needs access to the servers where that configuration is deployed | 15:01 |
| *** e0ne_ has joined #openstack-tc | 15:02 | |
| fungi | we try to make those things debuggable without access to the servers for the most part | 15:02 |
| zbr|rover | fungi: not root, just ability to access logs. | 15:02 |
| fungi | on rare occasions where it doesn't work as expected, it's fine to reach out to a smaller group of people with root access to dig in logs which may have lots of sensitive data, credentials, pii, et cetera | 15:03 |
| *** e0ne has quit IRC | 15:05 | |
| ttx | If anything, that puts us in a virtuous circle of fixing those corner cases. Like if a project-level config change crashes servers. The more that separation works (project consumption vs. root services), the better | 15:05 |
| zbr|rover | ttx: i think nobody argues against separation. | 15:19 |
| fungi | as for who gets added to project-config-core that's probably technically my responsibility as tact sig chair, though i would consult with clarkb since it's sort of shared jurisdiction with opendev at the moment (the central opendev content of that repo is planned to move to the opendev/project-config repo eventually leaving just openstack-specific configuration in openstack/project-config, but that | 15:23 |
| fungi | transition will be fiddly) | 15:23 |
| fungi | it's not just that repo though, a lot of openstack-specific job configuration also resides in openstack/openstack-zuul-jobs | 15:28 |
| gmann | I can surly help on jobs side though I am not so familiar on other infra activities. | 15:31 |
| *** rpittau is now known as rpittau|afk | 15:38 | |
| *** dklyle has joined #openstack-tc | 15:48 | |
| fungi | that would help for sure. the bulk of the activity there is split between changes to central trusted job definition/instantiation, and changes creating new projects | 15:59 |
| *** gibi is now known as gibi_pto | 16:35 | |
| *** chandankumar is now known as raukadah | 16:53 | |
| *** e0ne_ has quit IRC | 17:54 | |
| *** ralonsoh has quit IRC | 17:57 | |
| *** andrewbonney has quit IRC | 18:33 | |
| *** e0ne has joined #openstack-tc | 19:36 | |
| *** e0ne has quit IRC | 20:10 | |
| *** e0ne has joined #openstack-tc | 20:22 | |
| *** e0ne has quit IRC | 20:40 | |
| *** slaweq has quit IRC | 20:54 | |
| *** e0ne has joined #openstack-tc | 21:36 | |
| *** e0ne has quit IRC | 21:41 | |
| *** dklyle has quit IRC | 23:50 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!