#meeting-3:fedoraproject.org log

<@james:fedora.im>

19:01:34

!startmeeting Fedora Infrastructure Ops Daily Standup Meeting

<@meetbot:fedora.im>

19:03:34

Meeting started at 2025-06-16 19:01:34 UTC

<@meetbot:fedora.im>

19:03:34

The Meeting name is 'Fedora Infrastructure Ops Daily Standup Meeting'

<@nirik:matrix.scrye.com>

19:03:45

morning

<@james:fedora.im>

19:03:48

Hey

<@phsmoura:fedora.im>

19:03:51

hello

<@james:fedora.im>

19:04:09

!topic Tickets needing review

<@james:fedora.im>

19:04:13

!info https://pagure.io/fedora-infrastructure/issues?status=Open&priority=1

<@james:fedora.im>

19:04:20

!ticket 12601

<@zodbot:fedora.im>

19:04:20

● **Last Updated:** Never

<@zodbot:fedora.im>

19:04:20

**fedora-infrastructure #12601** (https://pagure.io/fedora-infrastructure/issue/12601):**Decommission provisioning.fp.org**

<@zodbot:fedora.im>

19:04:20

● **Assignee:** Not Assigned

<@zodbot:fedora.im>

19:04:20

<@zodbot:fedora.im>

19:04:20

● **Opened:** 56 minutes ago by pbrobinson

<@nirik:matrix.scrye.com>

19:04:32

med med ops?

<@nirik:matrix.scrye.com>

19:04:42

should be pretty easy so perhaps med low?

<@james:fedora.im>

19:04:54

Yeh, it looks like we just turn it off and walk away?

<@james:fedora.im>

19:05:11

I'm fine with med though because openshift scares me :)

<@nirik:matrix.scrye.com>

19:05:14

delete it in openshift and remove stuff in ansible, and add a redirect at the end

<@nirik:matrix.scrye.com>

19:05:38

or we could do the lazy way...

<@nirik:matrix.scrye.com>

19:05:48

just ignore it for now, but not migrate it to the new datacenter. ;)

<@james:fedora.im>

19:06:13

Still need to do the ansible bits ;)

<@nirik:matrix.scrye.com>

19:06:20

but better to cleanly remove it

<@james:fedora.im>

19:06:56

No new releng tickets...

<@james:fedora.im>

19:06:57

!topic Planning, Upcoming work and Open floor

<@james:fedora.im>

19:07:24

!info Two weeks until DC move!

<@nirik:matrix.scrye.com>

19:07:42

!info 14 days until datacenter switcharoo week

<@nirik:matrix.scrye.com>

19:07:48

yep.

<@nirik:matrix.scrye.com>

19:08:01

So... I have some stuff you all could help with if you have cycles:

<@nirik:matrix.scrye.com>

19:08:37

phsmoura: would you be willing to install a proxy01.stg.rdu3 and proxy02.stg.rdu3 ? I can help with info needed... but it hopefully shouldn't be too bad.

<@phsmoura:fedora.im>

19:08:39

I finished to reinstall the proxies should we close the issue? also opened a PR adding tmpwatch back to ansible

<@phsmoura:fedora.im>

19:08:52

sure

<@nirik:matrix.scrye.com>

19:08:56

ah, did tmpwatch get back in? cool.

<@james:fedora.im>

19:09:13

Is the fixed version in the official repos?

<@phsmoura:fedora.im>

19:09:15

seems so, their PR got merged

<@nirik:matrix.scrye.com>

19:10:07

James: you interested in setting up rabbitmq clusters in rdu3? :) If we do rhel9 we need to get rabbitmq-server from the centos messaging sig... we could also look at rhel10 for it I guess since it's an anoying one to upgrade.

<@james:fedora.im>

19:10:56

I'm fine to try anything ... was expecting to take wiki ;) ... and maybe some of the easier openshift extras ... I can certainly try rabbitmq though.

<@nirik:matrix.scrye.com>

19:11:08

epel10 right now lacks nagios and zabbix_client and collectd... but otherwise things all see there.

<@nirik:matrix.scrye.com>

19:11:58

so for this week my plan is: get openshift setup the rest of the way and stg openshift (but that needs staging proxies), then get fcos and some projects that don't have database or nfs to move.

<@nirik:matrix.scrye.com>

19:12:11

also want to install builders and power10/arm stuff

<@phsmoura:fedora.im>

19:12:13

ah sorry.. the links is redirecting to an issue now the PR page is 404 :/ mt mistake

<@james:fedora.im>

19:12:22

Can/should we run without any of those? Do we want to poke people, or try to get builds into our infra repo?

<@phsmoura:fedora.im>

19:12:23

ah sorry.. the links is redirecting to an issue now the PR page is 404 :/ my mistake

<@phsmoura:fedora.im>

19:12:38

ah sorry.. the link is redirecting to an issue now the PR page is 404 :/ my mistake

<@nirik:matrix.scrye.com>

19:12:43

anything with databases or nfs we can't easily move early. It will need to be duing switch week.

<@nirik:matrix.scrye.com>

19:13:08

colllectd is just my lack of time. I am not sure on the other two...

<@james:fedora.im>

19:13:28

I wasn't sure if we could install the wiki, but have no data ... and then "just" switch the data during the 30th week.

<@nirik:matrix.scrye.com>

19:13:52

yeah, we can... and probibly should next week...

<@nirik:matrix.scrye.com>

19:14:04

but I suppose we could do that part early.

<@james:fedora.im>

19:14:06

Okay, I'll look at build of collectd and pinging people for the monitoring.

<@nirik:matrix.scrye.com>

19:14:48

I didn't want to get too far into adjusting for rhel10 since we are so low on time, but for rabbitmq and databases it might be worth doing to avoid upgrades later.

<@james:fedora.im>

19:14:54

We have 9ish work days, so I'm fine with whatever order you'd prefer

<@james:fedora.im>

19:15:40

Yeh, I'll look at el10 for it this week ... then if it's not there we can go with plan B of el9 next week.

<@nirik:matrix.scrye.com>

19:15:50

well, we can see how easy it is to just build nagios/zabbix/collecd in epel10-infra for now I guess.

<@james:fedora.im>

19:16:00

!info Thursday is a federal holiday in the US (19th June).

<@nirik:matrix.scrye.com>

19:16:05

but there's also a ticket about adding epel10-infra tags

<@nirik:matrix.scrye.com>

19:17:05

phsmoura: so on those stg proxies, should just be add inventory and host_vars for them... ip's are the same but 10.16 instead of 10.3... and vmhost's should use .rdu3. versions instead of .iad2.

<@phsmoura:fedora.im>

19:18:24

ok, should run ansible after updating that?

<@james:fedora.im>

19:18:26

!releng 12765

<@zodbot:fedora.im>

19:18:27

**releng #12765** (https://pagure.io/releng/issue/12765):**make epel10-infra tags in koji**

<@zodbot:fedora.im>

19:18:27

<@zodbot:fedora.im>

19:18:27

● **Opened:** 2 weeks ago by kevin

<@zodbot:fedora.im>

19:18:27

● **Last Updated:** 2 weeks ago

<@zodbot:fedora.im>

19:18:27

● **Assignee:** patrikp

<@nirik:matrix.scrye.com>

19:19:05

well, adding... we do not want to remove the iad2 ones, we want to add new rdu3 ones. ;) but yeah, after adding those you should be able to run ansible to deploy them...

<@nirik:matrix.scrye.com>

19:19:25

same for rabbitmq... and any other rdu3 vm we are deploying.

<@james:fedora.im>

19:19:30

Having never added tags before, that doesn't seem like a real blocker.

<@nirik:matrix.scrye.com>

19:19:49

it shouldn't be hard, just I haven't had time to do it. ;)

<@nirik:matrix.scrye.com>

19:20:56

well, koji has it's own permissions... so koji admins are seperate from any fas group

<@nirik:matrix.scrye.com>

19:21:02

but yeah.

<@james:fedora.im>

19:21:05

ahh, I guess not then ;)

<@nirik:matrix.scrye.com>

19:21:32

I actually have an idea of how to make koji use groups, but it's... down on the todo list

<@nirik:matrix.scrye.com>

19:21:47

anyhow, any other questions?

<@phsmoura:fedora.im>

19:22:20

Im good

<@james:fedora.im>

19:22:54

No real questions ... did notice you were doing openshift install things over the weekend. Do you think we are good for time?

<@nirik:matrix.scrye.com>

19:23:41

hard to say... I hope so? but I am going to try and get lots done this week... better to mostly be ready at the end than scrambling.

<@james:fedora.im>

19:25:10

Anyway, on that cheery note I'll give y'all 5 minutes

<@james:fedora.im>

19:25:15

!endmeeting