<@pboy:fedora.im>
18:00:20
!startmeeting fedora-server
<@meetbot:fedora.im>
18:00:21
Meeting started at 2024-03-06 18:00:20 UTC
<@meetbot:fedora.im>
18:00:21
The Meeting name is 'fedora-server'
<@pboy:fedora.im>
18:00:30
!topic Welcome / roll call
<@pboy:fedora.im>
18:00:38
Welcome to our Server WG meeting today!
„Same procedure as every year“ We'll give a few minutes for folks to show up
<@pboy:fedora.im>
18:00:52
I’ll post the agenda in about 2 minutes.
Sorry for the late Agenda. We had a breakdown of our house heetring, which caused some headaches, brought a lot of discomfort and cost even more euros. Fortunately, everything is back again.
<@eseyman:fedora.im>
18:01:30
Ah, that sucks...
<@humaton:fedora.im>
18:01:34
hello
<@eseyman:fedora.im>
18:01:34
Hey, everybody
<@conan_kudo:matrix.org>
18:02:22
!hi
<@zodbot:fedora.im>
18:02:23
Neal Gompa (ngompa) - he / him / his
<@carlwgeorge:matrix.org>
18:02:48
!hi
<@zodbot:fedora.im>
18:02:50
Carl George (carlwgeorge) - he / him / his
<@eseyman:fedora.im>
18:03:14
!hi
<@zodbot:fedora.im>
18:03:15
Emmanuel Seyman (eseyman) - he / him / his
<@jwhimpel:matrix.org>
18:03:40
!hi
<@zodbot:fedora.im>
18:03:42
No Fedora Accounts users have the @jwhimpel:matrix.org Matrix Account defined
<@cooltshirtguy:matrix.org>
18:04:04
!hi
<@zodbot:fedora.im>
18:04:07
No Fedora Accounts users have the @cooltshirtguy:matrix.org Matrix Account defined
<@cooltshirtguy:matrix.org>
18:04:32
:P
<@pboy:fedora.im>
18:05:03
Matrix/zodbot seems a little bitchy today.
<@pboy:fedora.im>
18:05:49
Nevertheless, the quorum has been reached.
<@pboy:fedora.im>
18:06:13
Zodbot will hopefully come to his senses
<@pboy:fedora.im>
18:06:26
!topic Agenda
<@pboy:fedora.im>
18:06:35
!info Follow-up actions & announcements
<@pboy:fedora.im>
18:06:44
!info Review installation media
<@pboy:fedora.im>
18:06:52
!info LVM2 configuration & ARM SBC installation
<@pboy:fedora.im>
18:07:00
!info F40 Chang Set items requiring our attention
<@pboy:fedora.im>
18:07:07
!info Testing F40
<@pboy:fedora.im>
18:07:18
info Our "story" for F40
<@pboy:fedora.im>
18:07:25
!info Open Floor
<@pboy:fedora.im>
18:07:41
Anything to add?
<@pboy:fedora.im>
18:08:11
OK Let's start
<@pboy:fedora.im>
18:08:22
!topic 1. Follow-up actions & announcements
<@pboy:fedora.im>
18:08:32
At the moment, there is nothing new here.
<@pboy:fedora.im>
18:08:39
Just one thing:
<@pboy:fedora.im>
18:08:50
There is a question from the EPEL team if we can move our meeting forward to 16:00 URC so that EPEL can occupy the space at 18:00 UTC.
<@pboy:fedora.im>
18:09:00
What do you think?
For me it is doable, but difficult. It conflicts with our family dinner and the kids' bedtime routine.
But if I'm the only one, it doesn't have to fail because of me.
I'm a bit reluctant. We can only do this if it really is perfectly OK for each of us!
<@conan_kudo:matrix.org>
18:09:20
I would very much prefer this as it's been hard to come to these because of all the meeting conflicts
<@cooltshirtguy:matrix.org>
18:09:37
I'm ok with moving it.
<@conan_kudo:matrix.org>
18:09:50
I had just managed to free it up and then EPEL needed to move, so as a member of the EPEL steering committee and this, I would like it to move so I can attend both
<@carlwgeorge:matrix.org>
18:10:11
small extra note, the reason for 1600 and not 1700 is because the epel steering committee meeting pins to utc and doesn't change with summer time
<@carlwgeorge:matrix.org>
18:10:29
so i think that would work out to 1600/1700 for this meeting, and 1800 for epel
<@carlwgeorge:matrix.org>
18:11:24
alternatively, this meeting could do 1700 and pin to utc
<@eseyman:fedora.im>
18:11:25
doable but difficult for me
<@rcallicotte:fedora.im>
18:11:51
!hi
<@zodbot:fedora.im>
18:11:52
Robby Callicotte (rcallicotte) - he / him / his
<@eseyman:fedora.im>
18:11:53
can't we just meet in a "Fedora Meeting 1" channel ?
<@carlwgeorge:matrix.org>
18:12:20
epel steering already moved to that channel, but that doesn't help attendees of both meetings
<@conan_kudo:matrix.org>
18:12:21
the idea is to avoid double booking
<@eseyman:fedora.im>
18:12:33
True
<@pboy:fedora.im>
18:12:46
That's the other question. Can Epel use the other room as last week and all the time before?
<@eseyman:fedora.im>
18:12:49
Let's try it and see if we can make it work
<@carlwgeorge:matrix.org>
18:12:54
fyi, the server wg docs do actually say this meeting is supposed to be at 1700 utc and doesn't mention shifting with summertime. i only found out at the last meeting that it has been shifting to 1800 half the year.
<@pboy:fedora.im>
18:13:12
Double booking may be sometime unavoicable
<@pboy:fedora.im>
18:13:35
jwhimpel? jednorozec?
<@carlwgeorge:matrix.org>
18:13:36
epel already moved to #meeting-1:fedoraproject.org
<@humaton:fedora.im>
18:13:57
I am ok with any time
<@jwhimpel:matrix.org>
18:14:17
There's also meeting-room-2 and meeting-room-3. Are they available for us?
<@carlwgeorge:matrix.org>
18:14:18
it's not a room conflict anymore, the concern is for not double booking folks that want to attend both
<@pboy:fedora.im>
18:14:55
Hm, I suppose, it's Neil only?
<@jwhimpel:matrix.org>
18:15:20
If we need to move, I would rather make it earlier than later.
<@pboy:fedora.im>
18:15:31
And sorry, Neil, you didn't attend for more than a year now.
<@carlwgeorge:matrix.org>
18:16:23
i would be interesting in starting to attend, but won't be able to if it conflicts in time with epel
<@pboy:fedora.im>
18:16:30
The issue is: IoT's meeting is 16:00 UTC.
<@pboy:fedora.im>
18:16:49
I just noticed.
<@pboy:fedora.im>
18:17:06
Then we would have another double booking.
<@pboy:fedora.im>
18:17:56
Carl George: OK, that's an argument
<@carlwgeorge:matrix.org>
18:18:33
1700utc would work if it pins to utc time
<@pboy:fedora.im>
18:19:32
Pinning to UTC didn't work for many of us. Therefore we decided to shift.
<@humaton:fedora.im>
18:20:33
yeah pinning to UTC makes sense if you have no other obligations...
<@eseyman:fedora.im>
18:21:03
because those don't pin to UTC
<@humaton:fedora.im>
18:21:12
yup
<@jwhimpel:matrix.org>
18:21:25
It would not be my first preference to pin, but I could make it work.
<@carlwgeorge:matrix.org>
18:21:39
does iot pin to utc?
<@pboy:fedora.im>
18:22:14
As I see, our core team members would be able to shift. Problem is to find a good time.
<@carlwgeorge:matrix.org>
18:22:24
for epel at least it was our preference to pin to utc, because then you never have to ask whose summertime shift the shift is happening with. it's different in many countries.
<@pboy:fedora.im>
18:22:30
Don't know what IoT does
<@pboy:fedora.im>
18:23:33
I propose, I'll ask IoT and we discuss on mailing list and decide next meeting.
<@carlwgeorge:matrix.org>
18:23:51
checking their fedora calendar entries, iot also pins to utc
<@carlwgeorge:matrix.org>
18:23:59
server wg is the outlier
<@pboy:fedora.im>
18:24:21
Hm, all the other teams Im egaged did not pin :-)
<@conan_kudo:matrix.org>
18:24:39
fedora-kde and workstation pin to US/ET
<@conan_kudo:matrix.org>
18:24:53
it varies on the teams
<@carlwgeorge:matrix.org>
18:25:48
s/outlier/outlier in this narrow window/
<@pboy:fedora.im>
18:26:10
I see no objection to decide next meeting.
<@eseyman:fedora.im>
18:26:39
folks, we've spent 20 minutes on this. Let's make a decision and move on
<@pboy:fedora.im>
18:26:41
!agreed We decide next meeting about a shift of our meeting time.
<@humaton:fedora.im>
18:26:41
DST was never a thing world wide its just us the "west"
<@carlwgeorge:matrix.org>
18:26:52
pinning to a regions dst shifts is nice for people in that region, but double bad for everyone else
<@pboy:fedora.im>
18:27:01
Yes, we lost a lot of time here.
<@jwhimpel:matrix.org>
18:27:10
daylight savings starts here this weekend. Need to decide today for next meeting
<@carlwgeorge:matrix.org>
18:27:20
with pinning to utc, it's equivalent bad for everyone, just a singe conversion of utc to local
<@pboy:fedora.im>
18:28:03
jwhimpel: Next meeting is the same as today.
<@pboy:fedora.im>
18:28:45
Unless we decide otherwise on the mailing list.
<@jwhimpel:matrix.org>
18:28:50
Thx, I'll adjust my calendar.
<@pboy:fedora.im>
18:29:11
Let's go on. We lost 30 m ins!
<@pboy:fedora.im>
18:29:20
!topic 2. Review installation media
<@pboy:fedora.im>
18:29:27
!link https://pagure.io/fedora-server/issue/88
<@pboy:fedora.im>
18:29:34
We had a good discussion last meeting.
Anything new here?
<@humaton:fedora.im>
18:29:42
yes
<@pboy:fedora.im>
18:29:50
Good! Go omn
<@pboy:fedora.im>
18:30:00
omn -> on
<@humaton:fedora.im>
18:30:22
I didnt have much time to spend on this, but I have discovered that server config have some warnings in logs during compose
<@humaton:fedora.im>
18:30:39
2024-03-05 05:20:11 [WARNING ] Variant Server.aarch64 requires comps group ^server-product-environment which does not match anything in input comps file
<@humaton:fedora.im>
18:30:43
there is bunch of them
<@pboy:fedora.im>
18:30:54
Oh, sh....
<@humaton:fedora.im>
18:31:02
it means that we have some groups that we are installing for the DVD media that are not defined
<@humaton:fedora.im>
18:31:58
I have no Idea how to modify the netinstall image yet but getting there
<@humaton:fedora.im>
18:32:35
and continue from there
<@pboy:fedora.im>
18:32:38
Thank goodness we still get a working medium
<@pboy:fedora.im>
18:33:23
OK, can I have a look, too? I created the VM image, which may have the same issue?
<@carlwgeorge:matrix.org>
18:33:33
this is part of my point, dst doesn't start for peter in germany until march 31st, even though for many it already will have started by the time of the next many
<@humaton:fedora.im>
18:33:34
it was a warning from the beginning or we changed group definitions, but its warning so it just skip those groups
<@pboy:fedora.im>
18:34:07
OK, so the fix is just to remove them?
<@conan_kudo:matrix.org>
18:34:34
well, we need to know what they're for and adapt
<@humaton:fedora.im>
18:34:34
so the problem is here https://pagure.io/pungi-fedora/blob/main/f/variants-fedora.xml#_46
<@eseyman:fedora.im>
18:34:34
removing them won't hurt because they're not defined
<@carlwgeorge:matrix.org>
18:34:34
this is part of my point, dst doesn't start for peter in germany until march 31st, even though for many it already will have started by the time of the next meeting
<@humaton:fedora.im>
18:35:12
yeah I need to look into comps and see if they changed or were never there
<@humaton:fedora.im>
18:35:25
for those interested
<@humaton:fedora.im>
18:35:42
those group definitions of rpms are here https://pagure.io/fedora-comps
<@pboy:fedora.im>
18:36:22
When I created the VM image I noticed, that the server related comps groups need a reworking.
<@pboy:fedora.im>
18:37:16
I found, they are overlapping and partly contradictory
<@humaton:fedora.im>
18:37:27
current rawhide definitions are here https://pagure.io/fedora-comps/blob/main/f/comps-f41.xml.in
<@humaton:fedora.im>
18:37:48
note that it will be branched to -f42 once rawhide moves to f42
<@humaton:fedora.im>
18:38:07
so wee want to do changes in the f41
<@pboy:fedora.im>
18:38:32
OK, got it.
<@pboy:fedora.im>
18:38:54
I would like to have a look at the groups items.
<@eseyman:fedora.im>
18:38:57
jednorozec: you don't have a list of warnings on hand, do you?
<@humaton:fedora.im>
18:39:11
Emmanuel Seyman: I do
<@humaton:fedora.im>
18:39:20
let me paste it to some service
<@humaton:fedora.im>
18:39:56
https://paste.centos.org/view/74b07bb4
<@eseyman:fedora.im>
18:40:00
ah, that's great
<@pboy:fedora.im>
18:40:04
Maybe an issue in fedora-server?
<@humaton:fedora.im>
18:41:04
there might be more warnings in the compose I got only so far in the log from latest rawhide
<@humaton:fedora.im>
18:41:18
for those interested this is global log from rawhide compose https://kojipkgs.fedoraproject.org/compose/rawhide/latest-Fedora-Rawhide/logs/global/pungi.global.log
<@humaton:fedora.im>
18:41:31
seraching for Server might have some results
<@eseyman:fedora.im>
18:41:33
Strange, I do see the groups in comps...
<@humaton:fedora.im>
18:41:51
they might be wrongly grouped or something
<@humaton:fedora.im>
18:42:04
there are groups and group in the comps file
<@conan_kudo:matrix.org>
18:42:11
yeah tracking it as a fedora-server issue would help
<@eseyman:fedora.im>
18:42:15
Let's create an issue in pagure and discuss there
<@humaton:fedora.im>
18:42:21
that is all i have like I said not much time to play with it
<@humaton:fedora.im>
18:42:34
sure let me do that
<@pboy:fedora.im>
18:42:48
I think next we should check the compose groups and see how up to date they are. I don't know if other media also use some of these groups.
<@pboy:fedora.im>
18:43:24
As I remember from my VM prozess, there are some potential issues.
<@pboy:fedora.im>
18:45:14
Proposal !agreed Sever WG will check the current compose groups in issues on fedora-server repo.
<@humaton:fedora.im>
18:45:39
https://pagure.io/fedora-server/issue/130
<@pboy:fedora.im>
18:46:16
jednorozec: ++1
<@zodbot:fedora.im>
18:46:18
pboy gave a cookie to humaton. They now have 54 cookies, 5 of which were obtained in the Fedora 39 release cycle
<@eseyman:fedora.im>
18:46:58
thank you, jednorozec
<@humaton:fedora.im>
18:47:09
np that is all have for now
<@pboy:fedora.im>
18:47:21
Any comment about the proposal?
<@cooltshirtguy:matrix.org>
18:47:44
nope
<@pboy:fedora.im>
18:47:51
!agreed Sever WG will check the current compose groups in issues on fedora-server repo.
<@pboy:fedora.im>
18:48:11
!action jednorozec and pboy will prepare the lists or links.
<@pboy:fedora.im>
18:48:33
!topic 3. LVM2 configuration & ARM SBC installation
<@pboy:fedora.im>
18:48:41
!link https://bugzilla.redhat.com/show_bug.cgi?id=2247872
<@pboy:fedora.im>
18:48:50
I couldn't test today's test branch so see, if it is resolved now.
<@pboy:fedora.im>
18:49:05
A problem may be, how to initialize the correct value at first startup.
#link https://bugzilla.redhat.com/show_bug.cgi?id=2247872#c19
<@pboy:fedora.im>
18:49:17
This issue is not accepted as release blocker. So it can happen that we release a kind of crap image for the 2nd time in a row. Not a joy.
Affected is our VM image and partly the aarch64 image, if you install directly (Raspberry Devices) and do not use am-image-installer.
<@pboy:fedora.im>
18:50:15
But LVM group seems to be working hard on it.
<@pboy:fedora.im>
18:50:48
That's just for information, we should skip to F40 change set.
<@pboy:fedora.im>
18:51:00
!topic 4. F40 Chang Set items requiring our attention
<@pboy:fedora.im>
18:51:22
Emmanuel Seyman: It's yours :-)
<@eseyman:fedora.im>
18:52:24
I went through the list of accepted F40 Changes looking for any ones that might be of concern/interest to us
<@eseyman:fedora.im>
18:52:35
I published the list on the mailing-list
<@pboy:fedora.im>
18:53:05
Yeah, sorry I missed to publish the link
<@eseyman:fedora.im>
18:53:30
https://lists.fedoraproject.org/archives/list/server@lists.fedoraproject.org/message/CNH57V7LT4GMQN676XFDJWQ6U46BGDMK/
<@jwhimpel:matrix.org>
18:54:34
This one: * Enable IPv4 Addresses Conflict Detection by Default -- It will affect startup time minimally, but I think we can live with it.
<@eseyman:fedora.im>
18:54:35
some stuff I'll be testing (MariaDB, PostgreSQL, 389 DS, ...) but I don't have the bandwidth to go through the entire list
<@eseyman:fedora.im>
18:56:33
I erred on the side of caution when deciding what to keep and what to drop
<@humaton:fedora.im>
18:57:41
I have some AMD HW so can test ROCm
<@jwhimpel:matrix.org>
18:59:50
I have a F40 system running several VM's successfully.
<@cooltshirtguy:matrix.org>
19:00:08
do we have testing procedures with this stuff? I've been around for a while but never had info on this kind of stuff.
<@eseyman:fedora.im>
19:00:49
not really no
<@eseyman:fedora.im>
19:01:16
(which is why I'm limiting myself to what I already know)
<@cooltshirtguy:matrix.org>
19:02:13
ahh ok
<@humaton:fedora.im>
19:02:43
so maybe we could put something together
<@humaton:fedora.im>
19:03:09
like test minimal functionality of web servers,db and such
<@humaton:fedora.im>
19:03:20
if they are part of the changes
<@humaton:fedora.im>
19:03:44
like now when mysql is here again and it is not packaged nicely right now
<@cooltshirtguy:matrix.org>
19:04:14
I like that idea
<@pboy:fedora.im>
19:04:52
Yes, that is on our agenda, too, for a long time.
<@pboy:fedora.im>
19:04:55
!link https://pagure.io/fedora-server/issue/61
<@eseyman:fedora.im>
19:05:17
Let's kill two birds with one stone, then
<@pboy:fedora.im>
19:05:27
Good idea!
<@jwhimpel:matrix.org>
19:05:39
Sorry, I have to run now. Bye
<@pboy:fedora.im>
19:05:57
Thanks jwhimpel, bye
<@pboy:fedora.im>
19:06:20
What is the result of this topic? how do we proceed further?
<@pboy:fedora.im>
19:06:33
(our time is up)
<@eseyman:fedora.im>
19:07:06
We need to assign one or more people to each change, ask them to test and give us their testing criteria when they're finished
<@pboy:fedora.im>
19:08:31
Emmanuel Seyman: What d you think, would it be helpful, if I make a table in an issue? It may be easier to access as mailing list.
<@eseyman:fedora.im>
19:09:22
I don't know. I actually think we should use the list more than we already do...
<@pboy:fedora.im>
19:09:38
OK, good for me, too.
<@pboy:fedora.im>
19:09:50
So let's continue on mailing list.
<@eseyman:fedora.im>
19:10:23
and now dinner...
<@eseyman:fedora.im>
19:10:39
see you at the next meeting, folks
<@pboy:fedora.im>
19:10:44
!agreed Server WG will continue to discuss this topic on mailing list.
<@eseyman:fedora.im>
19:11:02
Peter Boy: I hope to reply to your email by Friday
<@cooltshirtguy:matrix.org>
19:11:28
thanks everyone
<@pboy:fedora.im>
19:11:32
Emmanuel Seyman: Would be fine. Sunday 10. is the last day
<@pboy:fedora.im>
19:11:52
Thanks to all for coming. See you in 14 days!
<@pboy:fedora.im>
19:12:04
!endmeeting