centos-hyperscale-sig
LOGS
<@conan_kudo:matrix.org>
16:00:43
!startmeeting CentOS Hyperscale SIG
<@meetbot:fedora.im>
16:00:45
Meeting started at 2025-11-05 16:00:43 UTC
<@meetbot:fedora.im>
16:00:45
The Meeting name is 'CentOS Hyperscale SIG'
<@conan_kudo:matrix.org>
16:00:52
!topic Roll call
<@conan_kudo:matrix.org>
16:01:02
!hi
<@zodbot:fedora.im>
16:01:06
Neal Gompa (ngompa) - he / him / his
<@rcolebaugh:matrix.org>
16:01:42
!hi
<@salimma:fedora.im>
16:01:43
!hi
<@zodbot:fedora.im>
16:01:45
Michel Lind (salimma) - he / him / his
<@noxiouz:fedora.im>
16:01:46
!hi
<@zodbot:fedora.im>
16:01:48
Raymond Colebaugh (rcolebaugh) - he / him / his
<@zodbot:fedora.im>
16:01:48
Anton Tiurin (noxiouz)
<@davide:cavalca.name>
16:02:03
!hi
<@zodbot:fedora.im>
16:02:05
Davide Cavalca (dcavalca) - he / him / his
<@james:reilly.asia>
16:03:09
!hi
<@zodbot:fedora.im>
16:03:12
James Reilly (jreilly1821) - he / him / his
<@conan_kudo:matrix.org>
16:05:07
seems like that's everyone
<@conan_kudo:matrix.org>
16:05:12
so let's get started with...
<@conan_kudo:matrix.org>
16:05:18
!topic Followups
<@conan_kudo:matrix.org>
16:06:16
anyone got anything to follow up from last meeting?
<@salimma:fedora.im>
16:07:02
did we finish onboarding James?
<@salimma:fedora.im>
16:07:27
now that hopefully travel/DST/etc. shenanigans are mostly over
<@conan_kudo:matrix.org>
16:07:31
good question
<@davide:cavalca.name>
16:07:50
Yep should be all set
<@conan_kudo:matrix.org>
16:08:15
excellent
<@conan_kudo:matrix.org>
16:08:52
I did notice that james hasn't signed the FPCA
<@conan_kudo:matrix.org>
16:09:15
james: can you take care of that in FAS?
<@conan_kudo:matrix.org>
16:09:41
you should find it here: https://accounts.fedoraproject.org/user/jreilly1821/settings/agreements/
<@conan_kudo:matrix.org>
16:11:10
anything else from anyone?
<@salimma:fedora.im>
16:11:54
not for follow ups, I have some for later
<@conan_kudo:matrix.org>
16:13:10
okay, since nobody has anything, let's move onto...
<@conan_kudo:matrix.org>
16:13:18
!topic Announcements
<@conan_kudo:matrix.org>
16:13:25
I've got one...
<@conan_kudo:matrix.org>
16:13:40
!info Rebase to 6.17 is in progress for Hyperscale
<@conan_kudo:matrix.org>
16:13:47
blech
<@conan_kudo:matrix.org>
16:14:02
!info Rebase to Linux kernel 6.17 is in progress for Hyperscale
<@conan_kudo:matrix.org>
16:14:35
I am starting on this work today
<@conan_kudo:matrix.org>
16:15:24
anyone have any other announcements?
<@salimma:fedora.im>
16:15:47
me
<@salimma:fedora.im>
16:16:09
!info kexec-tools for hs+fb.el9 rebased and tagged
<@salimma:fedora.im>
16:16:37
we can discuss in open floor later whether we should help upstream sort this package out, because it pulls in NM unnecessarily (and that's what we worked around by loosening some deps)
<@salimma:fedora.im>
16:16:42
and one more
<@salimma:fedora.im>
16:17:28
!info wprof, a BPF profiler, will be built in Hyperscale sometime later today. once blazesym is fixed in Fedora we'll try moving it there and unbundle the blazesym copy it carries
<@conan_kudo:matrix.org>
16:17:44
sweet
<@conan_kudo:matrix.org>
16:18:10
is there a website or something about it?
<@salimma:fedora.im>
16:18:22
https://github.com/anakryiko/wprof
<@conan_kudo:matrix.org>
16:18:49
neato
<@salimma:fedora.im>
16:18:51
it's been mostly used by kernel devs, I had to ask Andrii to write blurbs for about and desc to even put something in the package :)
<@salimma:fedora.im>
16:18:55
I am sure he will blog about it soon
<@conan_kudo:matrix.org>
16:19:02
cool
<@conan_kudo:matrix.org>
16:19:22
ah blazesym is a rust-based library
<@conan_kudo:matrix.org>
16:19:49
oh hey it supports breakpad tracing
<@salimma:fedora.im>
16:19:50
yeah, I packaged blazesym already, and blazesym_c, but they have been languishing since nobody uses it
<@conan_kudo:matrix.org>
16:19:53
that's pretty cool
<@salimma:fedora.im>
16:20:54
I don't have anything else to bring up
<@conan_kudo:matrix.org>
16:21:17
anyone else have anything?
<@conan_kudo:matrix.org>
16:22:52
if not, let's move onto...
<@conan_kudo:matrix.org>
16:23:01
!topic Tickets
<@conan_kudo:matrix.org>
16:23:27
I don't think we have any new tickets here, so we can move on from this
<@conan_kudo:matrix.org>
16:23:52
so let's move onto...
<@conan_kudo:matrix.org>
16:24:00
!topic Membership
<@conan_kudo:matrix.org>
16:24:15
I don't think there are any new membership requests, but...
<@conan_kudo:matrix.org>
16:24:30
we need to start pruning inactive members
<@conan_kudo:matrix.org>
16:24:48
it's nice being the biggest SIG, but being full of ghosts isn't great either...
<@conan_kudo:matrix.org>
16:25:43
Davide Cavalca: we might want to put this on the todo for our meetup at Connect
<@davide:cavalca.name>
16:26:17
I think we just need to write up and announce a policy
<@davide:cavalca.name>
16:26:28
And then execute it
<@salimma:fedora.im>
16:26:53
yeah - I think I have a backlog task to write an activity script
<@davide:cavalca.name>
16:26:57
We can discuss it at the meetup but I don't think this necessarily needs to be sorted out in person
<@salimma:fedora.im>
16:27:14
but we should probably just start the policy write up in parallel
<@conan_kudo:matrix.org>
16:27:15
if we can get it done sooner, that'd be great
<@salimma:fedora.im>
16:27:22
and if it turns out we'll be pruning too many people we can adjust
<@james:reilly.asia>
16:27:39
180 days of no interaction or something?
<@conan_kudo:matrix.org>
16:28:02
something like that indeed
<@salimma:fedora.im>
16:29:35
yeah, I think 6 months is nice as it's one minor release
<@salimma:fedora.im>
16:30:06
and I think we want to promote the actually useful members we retire to emeritus status, but for those that never do anything we can just drop them
<@conan_kudo:matrix.org>
16:31:27
sure
<@conan_kudo:matrix.org>
16:31:40
we'll need to discuss a way to represent that, but I don't see a problem with that
<@salimma:fedora.im>
16:33:41
who wants to volunteer to wordsmith this :)
<@conan_kudo:matrix.org>
16:33:56
I can't, I'm swamped
<@davide:cavalca.name>
16:35:19
This is a can of worms that's IMO not worth the trouble. Let's just have a cutoff and keep a record of former members in the docs.
<@salimma:fedora.im>
16:36:02
yeah, true
<@conan_kudo:matrix.org>
16:40:48
okay, I think in the interest of ending on time, let's punt this for later
<@conan_kudo:matrix.org>
16:41:02
we do have the meetup in the evening if we want to discuss it then too
<@conan_kudo:matrix.org>
16:41:26
so let's move to...
<@conan_kudo:matrix.org>
16:41:30
!topic Miscellaneous / Open Floor
<@conan_kudo:matrix.org>
16:41:46
anything anyone want to bring up?
<@salimma:fedora.im>
16:42:21
how likely is it that we can get RH to fix kexec-tools' packaging? anyone knows the maintainers?
<@davide:cavalca.name>
16:42:38
I filed PRs a while ago to not much success
<@davide:cavalca.name>
16:42:46
On the Fedora side I mean
<@salimma:fedora.im>
16:43:54
sigh
<@salimma:fedora.im>
16:44:09
so probably not worth putting it in proposed updates and shame upstream then
<@conan_kudo:matrix.org>
16:44:11
pretty likely with the right sticks
<@salimma:fedora.im>
16:44:15
since who has time for that
<@conan_kudo:matrix.org>
16:44:30
we can always ask folks like Troy Dawson for help
<@conan_kudo:matrix.org>
16:45:02
I do have decent success with packaging fixes, at least once they're in fedora
<@conan_kudo:matrix.org>
16:45:15
but also, I often just merge them in fedora if nobody responds for a while
<@salimma:fedora.im>
16:45:45
yeah we probably should just go that route
<@salimma:fedora.im>
16:45:53
or even, spite hat on, nonresponsive maintainer the persons
<@conan_kudo:matrix.org>
16:46:19
well we have that new stalled process thing
<@salimma:fedora.im>
16:46:24
for clarity, no, I was not serious about that
<@salimma:fedora.im>
16:46:39
yeah I think forcing the merge of the PR via that process is not the worst option
<@salimma:fedora.im>
16:46:50
it's not urgent enough to abuse provenpackager privileges
<@salimma:fedora.im>
16:47:04
but also we want this on the record so FESCo know some critical packages are not being fixed by their maintainers
<@conan_kudo:matrix.org>
16:47:12
one package that makes my head hurt is the iptables package
<@conan_kudo:matrix.org>
16:48:25
anyway, we have pathways to resolve these issues
<@conan_kudo:matrix.org>
16:48:35
we're kind of bad at using them though
<@salimma:fedora.im>
16:48:40
I want to rip it out of my system knowing how bad it is
<@salimma:fedora.im>
16:48:46
but looks like firewalld and ufw both require it too
<@salimma:fedora.im>
16:49:01
a lot of these pathways require a lot of attention
<@salimma:fedora.im>
16:49:08
which is good and bad, we don't want them to be easy to abuse
<@salimma:fedora.im>
16:49:12
but bad because everyone is overworked
<@conan_kudo:matrix.org>
16:49:15
yeah
<@salimma:fedora.im>
16:49:23
so we pile up tech debts instead
<@conan_kudo:matrix.org>
16:49:32
honestly in most cases I just use my provenpackager powers to fix bugged packages
<@conan_kudo:matrix.org>
16:49:44
if nobody responds after a bit, I just take care of it
<@conan_kudo:matrix.org>
16:50:22
it's less painful that way
<@conan_kudo:matrix.org>
16:50:37
and honestly quicker
<@salimma:fedora.im>
16:52:50
yeah that might be the only way forward here that does not involve pulling teeth
<@conan_kudo:matrix.org>
16:54:01
if they don't care enough to respond in a week or two for PRs, they probably won't mind you fixing things
<@conan_kudo:matrix.org>
16:56:19
is there anything else we want to discuss?
<@salimma:fedora.im>
16:57:52
we can get... 3 mins back :)
<@conan_kudo:matrix.org>
16:57:59
yup
<@conan_kudo:matrix.org>
16:58:07
a minute saved is a minute earned!
<@conan_kudo:matrix.org>
16:58:17
!endmeeting