17:00:17 <pboy> #startmeeting fedora-server
17:00:17 <zodbot> Meeting started Wed Jun 15 17:00:17 2022 UTC.
17:00:17 <zodbot> This meeting is logged and archived in a public location.
17:00:17 <zodbot> The chair is pboy. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
17:00:17 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:00:17 <zodbot> The meeting name has been set to 'fedora-server'
17:00:26 <pboy> #topic Welcome / roll call
17:00:35 <pboy> Welcome to our Server WG IRC meeting today!
17:00:42 <pboy> „Same procedure as every year“ We'll give a few minutes for folks to show up
17:00:50 <pboy> Please, everybody who is lurking, say either .hello2 or .hello <fasname>
17:00:58 <pboy> I’ll post the agenda in a few minutes.
17:01:40 <cooltshirtguy> .hello2
17:01:41 <zodbot> cooltshirtguy: cooltshirtguy 'Jason Beard' <jas_beard@hotmail.com>
17:02:20 <pboy> Welcome cooltshirtguy
17:03:03 <cooltshirtguy> hello pboy
17:03:47 <salimma> .hi
17:03:48 <zodbot> salimma: salimma 'Michel Alexandre Salim' <michel@michel-slm.name>
17:04:07 <pboy> Welcome salimma!
17:04:57 <pboy> We seem to be in a small group today
17:05:38 <pboy> But we now meet the quorum
17:05:58 <pboy> I'll start with the agenda. Maybe some  will be late.
17:06:16 <pboy> #topic Agenda
17:06:24 <pboy> #link https://pagure.io/fedora-server/report/Meeting
17:06:32 <pboy> #info Follow up actions
17:06:39 <pboy> #info  Further processing of GPT as default partitioning switch
17:06:47 <pboy> #info  Test planning for Fedora 37
17:06:54 <pboy> #info How to proceed with Cockpit File Sharing module (continued)
17:07:01 <pboy> #info Open Floor
17:07:12 <pboy> Any additional topic / issue / comment ?
17:07:25 <pboy> You may get an overview of our ongoing and outstanding tasks at
17:07:32 <pboy> #link https://pagure.io/fedora-server/boards/Works%20in%20progress
17:08:40 <pboy> OK, no additions so far.
17:08:50 <pboy> #topic  Follow up actions
17:08:59 <pboy> #info Change proposal to add Server VM is still processing
17:09:06 <pboy> #link Proposal https://fedoraproject.org/wiki/Changes/Supplement-server-by-kvm-vm-image
17:09:16 <pboy> #link Discussion https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/thread/M2YQMVVUCFCV4MMOQ32UMSM5WBBVE2H7/
17:09:43 <pboy> Discussion is a bit exaggerated. Nothing so far
17:09:54 <pboy> #info There is now a Change Proposal regarding the default hostname configuration
17:10:02 <pboy> #link Proposal https://fedoraproject.org/wiki/Changes/FallbackHostname
17:10:13 <pboy> #link Discussion https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/thread/Y2TT6VZPGTD5UVGPA6PLNYW2BU4JOC77/
17:10:29 <pboy> Anyone for a break to read the Default hostname configuration discussion thread?
17:11:24 <cooltshirtguy> yes
17:12:01 <pboy> OK. we wait.
17:12:28 <pboy> I expect, the change will be accepted.
17:13:39 <pboy> As a follow up we then should adjust the naming of our volume group, I guess. I suppose, it's not that complicated.
17:14:43 <davide> .hello dcavalca
17:14:44 <zodbot> davide: dcavalca 'Davide Cavalca' <dcavalca@fb.com>
17:15:05 <pboy> Hello dcavalca! Welcome"
17:15:53 <pboy> dcavalca We have just a break to read the Default hostname change discussion on devel
17:16:14 <davide> catching up on that now, thanks
17:16:17 <mowest> .hello2
17:16:18 <zodbot> mowest: mowest 'Steve Daley' <mowest@vivaldi.net>
17:16:26 <eseyman> .hello2 eseyman
17:16:27 <zodbot> eseyman: eseyman 'Emmanuel Seyman' <emmanuel@seyman.fr>
17:16:41 <pboy> Welcome mowest and eseyman !
17:16:56 <eseyman> sorry, I'm late
17:17:14 <pboy> Better late then not at all. :-)
17:17:59 <pboy> We just have still a break to read the Default hostname configuration discussion thread
17:19:38 <cooltshirtguy> I'm ok with localhost or getting a hostname from dhcp. being a server the hostname will set to a standard via cloud-init or some kind of automation
17:19:43 <pboy> OK, are we ready? Any comment / idea to the hostname topic?
17:20:28 <pboy> Yes, I think it is in any case better as the current fixed "fedora".
17:20:48 <salimma> anyone has thoughts on the other suggestion that came up in the thread (using fedora-<uniquename>)?
17:20:50 <mowest> Agreed
17:20:53 <cooltshirtguy> agreed
17:20:56 <davide> I like the idea of leveraging systemd-hostnamed for this down the road, and just generate a unique fqdn
17:20:59 <eseyman> hostname from dhcp is my normal so I'm fine with that
17:21:02 <salimma> but yeah IMHO just localhost is strictly better than 'just fedora'
17:21:12 <davide> but in the meantime, localhost is also fine, and better than fedora
17:21:23 <salimma> look at that, everyone is saying the same thing :)
17:21:51 <mowest> I tend to change to hostnames of my servers, so a generic default like localhost or getting from dhcp seems good.
17:21:53 <eseyman> agree that localhost is better than fedora
17:22:08 <pboy> I think a unique default hostname would be fine, especially for desktops.
17:22:17 <pboy> But that#s for F38, I think.
17:23:14 <pboy> #agreed Server WG is content with the default hostname change proposal
17:23:40 <pboy> No additional new actions to follow up. Or did I miss an open action?
17:23:55 <mowest> +1
17:24:02 <salimma> not really, apart from testing once the change is implemented I guess?
17:24:23 <eseyman> per usual
17:24:30 <pboy> Yes, testing is another case.
17:24:41 <pboy> Well. let's proceed then.
17:24:52 <pboy> #topic Further processing of GPT as default partitioning switch
17:25:01 <pboy> We have currently no issue for this topic.
17:25:08 <pboy> But we have a discussion:
17:25:16 <pboy> #link https://lists.fedoraproject.org/archives/list/server@lists.fedoraproject.org/thread/XHFJ67E7PXR2WHMR2766TSM7VUWCO6PF/
17:25:32 <pboy> My intention is: Agreement on the further course of action
17:25:46 <pboy> Should we make break for reading?
17:26:14 <salimma> yes please
17:26:26 <pboy> OK 3 mins. :-)
17:26:37 <eseyman> document what works and what doesn't; promote Fedora Server keeping this in mind
17:27:50 <salimma> Right. documenting as soon as we can what's known not to work will help drive the implementation of "GPT by default" e.g. maybe it might have to be Workstation first, and/or fixing software RAID needs to be prioritized
17:29:15 <pboy> OK, I take that as a task to create a bugzilla entry for the uefi bot right now (it topic (b) )
17:29:16 <salimma> do we have any idea how prevalent software RAID is? this is 'traditional' software RAID, right, as opposed to LVM RAID?
17:29:22 <salimma> https://www.systutorials.com/docs/linux/man/7-lvmraid/
17:29:58 <eseyman> ah, I thought we were talking about LVM RAID all along
17:30:15 <mowest> Personally, in the home lab I use software raid with btfs on the data drives. I don't use RAID on my boot drive.
17:30:18 <pboy> salimma: It's a traditional raid, with a  vg ont top of it.
17:31:30 <pboy> That's the way Anaconda offers.
17:31:36 <pboy> As far as I know
17:32:26 <salimma> eseyman: yeah, I wanted to clarify because I'm assuming LVM RAID would work fine. that could be an option - we might want to look into if Anaconda can support LVM RAID or not
17:32:26 <cooltshirtguy> I personally only use RAID for data drives not boot drives. I might mirror the boot drive with LVM
17:32:56 <pboy> We don't know how many use SW raid. We have no data. I just know a lot of discussions and questions about sw raid.  It seems specifically used in private use cases.
17:33:16 <pboy> salima: good idea. I'll try it before I file a bug.
17:33:23 <salimma> right, for production use I'd expect hardware RAID to be more typical
17:33:53 <pboy> Yes, I think every server hardware has it onboard.
17:34:08 <Eighth_Doctor> Hey all
17:34:09 <Eighth_Doctor> .hello ngompa
17:34:10 <zodbot> Eighth_Doctor: ngompa 'Neal Gompa' <ngompa13@gmail.com>
17:34:29 <mowest> Hi Neal
17:34:44 <Eighth_Doctor> Hardware raid is often present even on consumer boards these days
17:35:48 <pboy> OK, we had   until now:
17:36:14 <pboy> #action pboy to add test of uefi sw raid with LVM RAID
17:37:15 <pboy> And I think, we agree to file a bug about uefi sw raid now and don't wait until the biosboot case is resolved?
17:37:29 <eseyman> yes, agree
17:37:31 <cooltshirtguy> agreed
17:37:34 <pboy> That's item (b) of the mail thread
17:37:54 <salimma> agreed
17:38:12 <pboy> action pboy files a bug about sw raid in uefi mode
17:38:44 <pboy> #action pboy files a bug about sw raid in uefi mode
17:39:20 <pboy> So, what about (c)?
17:39:57 <pboy> beyomd my typo is is about universal boot, to combine biosboot and efi partition
17:40:15 <pboy> Neil is the expert here, I think.
17:40:26 <pboy> Neal. .... sorry
17:40:37 <eseyman> this is out of my comfort zone; I'll abstain on c)
17:40:42 <eseyman> Eighth_Doctor: ^^
17:41:57 <eseyman> I think he's AFK
17:43:00 <pboy> I don't see a specific opinion. So let's postpone this until the current boat issues are resolved.
17:43:43 <pboy> Anything else on this topic?
17:43:50 <cmurf> for what it's worth the UEFI case for RAID is difficult because the installer can only do some suboptimal things
17:43:52 <pboy> 3
17:44:02 <pboy> 2
17:44:11 <pboy> 1
17:44:21 <cmurf> it's largely an unsolved problem without a service that takes responsibility for syncing multiple EFI System partitions
17:44:21 <pboy> #topic Test planning for Fedora 37
17:44:30 <pboy> We have currently no issue for this topic.
17:44:39 <pboy> Proposal: We should aim for a test *week* with Fedora QA and make this a permanent practice for the next releases.
17:44:46 <pboy> We would get something like:
17:44:54 <pboy> https://testdays.fedoraproject.org/events/132
17:45:31 <pboy> This would be a significant improvement over our approach to F36
17:45:39 <eseyman> huge +1 from me
17:45:45 <pboy> First time preparation is a lot of work and will be incomplete in a first shot. But we can carry it further to the next releases and improve it step by step.
17:45:51 <pboy> One advantage I see: We can offer our users a way to participate and also contribute their problem cases or important elements.
17:45:58 <pboy> And hopefully we will succeed in gaining one or the other active new member.
17:46:27 <mowest> +1 from me, having a specific time frame and clear goals for testing would help me focus my limited time for QA.
17:47:41 <cooltshirtguy> agree with mowest
17:47:56 <pboy> OK
17:48:07 <pboy> #agreed Server WG will strive for  a test week for F38 and follow up releases
17:48:28 <pboy> The plan would be:
17:48:36 <pboy> 1. Next meeting discussion of our Technical Specification
17:48:43 <pboy> 2. In 4 week’s meeting discussion of release criteria and a first collection of tests.
17:48:58 <pboy> 3. Contact Adam for implementation
17:49:21 <pboy> 3. Improve the test setup
17:49:32 <pboy> 3. -> 4.
17:50:19 <pboy> Any additional comment. / idea  here?
17:50:32 <pboy> 3
17:50:40 <pboy> 2
17:50:46 <pboy> 1
17:50:46 <eseyman> looks good to me
17:50:55 <pboy> #topic How to proceed with Cockpit File Sharing module
17:51:04 <pboy> #link https://pagure.io/fedora-server/issue/86
17:51:10 <pboy> As discussed I created 2 documentation stub regarding the NFS part that we needs to complete
17:51:26 <pboy> https://docs.stg.fedoraproject.org/en-US/fedora-server/service-filesharing-nfs-installation/
17:51:33 <pboy> https://docs.stg.fedoraproject.org/en-US/fedora-server/service-filesharing-nfs-administration-cockpit/
17:52:03 <pboy> It's difficult topic. Anyone any idea?
17:53:13 <mowest> I can only report that after my upgrade to F37 I saw the NFS file sharing in Cockpit, haven't tried setting it up through Cockpit. Not normally something I use, but I should try it to do QA on it.
17:54:19 <eseyman> my NFS needs are trivial and I don't need an admin module for Cockpit. I'll sit this one out
17:54:39 <pboy> Yeha, QA test is on thing. Another is to document, for what is is usable, and for what it is not.
17:54:40 <cooltshirtguy> personally i would setup NFS server through automation.
17:55:03 <cooltshirtguy> there's a push for cockpit for doing administrator tasks
17:55:11 <cooltshirtguy> though
17:55:46 <pboy> Well, Cockpit is one of our featured administration tools, beside CLI
17:56:31 <pboy> cooltshirtguy: Could you provide some ideas / hints we should include in an NFS installation article?
17:57:03 <eseyman> agreed but some things (apache and nginx also come to mind) can be configured 1000s of ways
17:57:26 <eseyman> so WebUI/GUI interfaces are hard to do
17:57:44 <pboy> eseyman: yes, but we should provide at leas one tested way of those thousends. :-)
17:58:03 <cooltshirtguy> I could give some. I haven't configured all the ways you can do NFS
17:58:42 <pboy> coolshirtguy: OK, let'do it on mailing list or so. So we will have a start.
17:58:57 <pboy> Well, we are running out of time
17:59:18 <pboy> Let's make a quick switch to open floor.
17:59:25 <pboy> #topic Open Floor
17:59:37 <pboy> So, any idea here?
18:00:17 <pboy> Obviously not.
18:00:23 <cooltshirtguy> hahaha
18:00:28 <pboy> Time is up, anyway
18:00:31 <mowest> We are end of time, so I'm good
18:00:44 <cooltshirtguy> all good
18:00:48 <eseyman> same here
18:00:51 <pboy> Yeah, was a good session today!
18:00:59 <pboy> Thanks everybody for comming.
18:01:13 <pboy> #endmeeting