<@pboy:fedora.im>
17:00:33
!startmeeting fedora-server
<@meetbot:fedora.im>
17:00:34
Meeting started at 2024-08-28 17:00:33 UTC
<@meetbot:fedora.im>
17:00:34
The Meeting name is 'fedora-server'
<@pboy:fedora.im>
17:00:46
!topic Welcome / roll call
<@pboy:fedora.im>
17:00:56
Let's wait a moment. I'll post the agenda in 2-3 minutes with a new business to deal with in addition to the agenda as sent earlier today.
<@pboy:fedora.im>
17:01:25
Sorry, copy&past error, no new business today :-)
<@jwhimpel:fedora.im>
17:02:03
!hi
<@zodbot:fedora.im>
17:02:04
John Himpel (jwhimpel)
<@eseyman:fedora.im>
17:03:34
!hello
<@zodbot:fedora.im>
17:03:35
Emmanuel Seyman (eseyman) - he / him / his
<@eseyman:fedora.im>
17:03:44
hello, folks
<@pboy:fedora.im>
17:04:06
Yeah, welcome everybody.
<@pboy:fedora.im>
17:04:46
I think, we should start, even if we are still missing at least 1 member.
<@pboy:fedora.im>
17:05:03
to reach our quorum
<@pboy:fedora.im>
17:05:17
!topic Agenda
<@pboy:fedora.im>
17:05:26
!info Follow-up actions & announcements
<@pboy:fedora.im>
17:05:35
!info Ansible assisted installation and configuration of NFS service
<@pboy:fedora.im>
17:05:47
!info Optimizing NFS service documentation
<@pboy:fedora.im>
17:05:57
!info Testing Release 41
<@pboy:fedora.im>
17:06:07
!info Open Floor
<@pboy:fedora.im>
17:06:29
Anythink to add?
<@eseyman:fedora.im>
17:06:37
I'm good
<@pboy:fedora.im>
17:06:57
OK, let's start
<@pboy:fedora.im>
17:07:08
!topic 1. Follow-up actions & announcements
<@pboy:fedora.im>
17:07:16
We have basically the same state as last week.
<@pboy:fedora.im>
17:07:28
So let's skip this for now.
<@pboy:fedora.im>
17:07:37
Any announcements?
<@pboy:fedora.im>
17:08:00
Obviously not.
<@eseyman:fedora.im>
17:08:07
No
<@pboy:fedora.im>
17:08:11
let's move on
<@pboy:fedora.im>
17:08:21
!topic 2. Ansible assisted installation and configuration of NFS service
<@pboy:fedora.im>
17:08:31
!link https://pagure.io/fedora-server/issue/138
<@pboy:fedora.im>
17:08:43
Continuing our discussion at last week's meeting about how we want to distribute the service specific Ansible playbooks.
<@pboy:fedora.im>
17:09:02
- we want to synchronize our manual installation guide, the Ansible doc/guide, and the Ansible process results (i.e. Ansible should create the same as the manual doc describes).
<@pboy:fedora.im>
17:09:02
We want to deal with
<@pboy:fedora.im>
17:09:02
- Discussion of the current first implementation
<@pboy:fedora.im>
17:09:02
- How do we proceed
<@pboy:fedora.im>
17:09:02
- What to do next
<@pboy:fedora.im>
17:09:02
<@pboy:fedora.im>
17:09:02
We may discuss
<@pboy:fedora.im>
17:09:02
<@pboy:fedora.im>
17:09:02
!link https://pagure.io/fedora-server/issue/138#comment-920322
<@pboy:fedora.im>
17:09:02
Our goal as discussed:
<@pboy:fedora.im>
17:11:01
So, I'm a bit unsure about the current structure of our Ansible repo. I would expect a playbook on the 1st level, something like HFS-service. Its now beyond task as main.yml.
<@pboy:fedora.im>
17:11:29
But maybe, that's currently not our most important question.
<@eseyman:fedora.im>
17:13:17
We don't currently have John's code in git, correct?
<@pboy:fedora.im>
17:13:34
I think we have, in stg.
<@jwhimpel:fedora.im>
17:14:12
My code is currently in the "stg" branch of our git repo.
<@eseyman:fedora.im>
17:14:40
ah, I didn't see that go through
<@eseyman:fedora.im>
17:15:35
I'll start working on that tonight
<@jwhimpel:fedora.im>
17:15:43
I'm currently working on implementing several of the suggestions Emmanuel suggested in his review
<@jwhimpel:fedora.im>
17:16:50
I'm trying to convert the client side to use systemd/mountd instead of /etc/fstab modifications.
<@jwhimpel:fedora.im>
17:17:42
I can mount the remote directories via cli "mount" command, but it fails with permission issues using systemd/mount role.
<@pboy:fedora.im>
17:18:54
Just a question, do we already have a short information / guidance, how I can use and test the code as far as we have it?
<@pboy:fedora.im>
17:21:38
On the practical side, I am an Ansible newbe. So maybe the ideal tester for a tutorial (All my knowledge about Ansible is theoretical, could write a comparison with other tools, but not practical).
<@jwhimpel:fedora.im>
17:22:18
I'll try to add usage instructions in the README.md in the next few days. It's mostly create a shell script to invoke the ansible role and modifying the variables found in the defaults/main.yml file to match your environment. Speaking of environment, you will need to define your hosts in the inventories directory. There you will also need to input your userid and password (encoded) so that you can connect to hosts.
<@jwhimpel:fedora.im>
17:24:43
A good place to start is: https://docs.ansible.com/ansible/latest/user_guide/index.html
<@jwhimpel:fedora.im>
17:27:42
I believe traditionally ansible role documentation is kept internal to the role in a README.md. But I think a Fedora Document outlining how to install/configure NFS server/client via ansible and cli should be made available in the Fedora Server section of Fedora documentation. Does that sound reasonable?
<@mowest:fedora.im>
17:28:19
!hello
<@zodbot:fedora.im>
17:28:20
Steve Daley (mowest)
<@eseyman:fedora.im>
17:28:32
Hi, Steve!
<@pboy:fedora.im>
17:28:41
John Himpel: Yes, is does sound reasonable in my opinion.
<@eseyman:fedora.im>
17:28:51
Agreed
<@mowest:fedora.im>
17:29:08
I'm a Ansible newbe as well so I would like to try testing too on a fresh install.
<@pboy:fedora.im>
17:29:39
mowest: So therfe afre 2 of us newbes :-)
<@mowest:fedora.im>
17:29:50
I agree with documentation idea as well.
<@jwhimpel:fedora.im>
17:30:20
Then I will suspend trying to implement systemd-mountd mounting and concentrate on README.md and creating content for the templates Peter has put in our repo.
<@pboy:fedora.im>
17:31:17
John Himpel: I think that is a good idea! So we could start to test / use it and could get a better understanding what we are achieving.
<@jwhimpel:fedora.im>
17:32:40
Question for the group: Do we want a static mount or do we want automount?
<@pboy:fedora.im>
17:33:37
Hm, considering about criteria to decide between the two.
<@pboy:fedora.im>
17:34:20
May it's a matter of performance? or the long-living of usage?
<@jwhimpel:fedora.im>
17:35:27
it doesn't have to be an either/or decision. In a round of revisions, it would be possible to chose either when running the ansible role (or cli)
<@jwhimpel:fedora.im>
17:35:56
s/round/ subsequent round/
<@pboy:fedora.im>
17:36:22
OK, so let's see and decide later.
<@pboy:fedora.im>
17:36:31
Other question of mine:
<@pboy:fedora.im>
17:36:43
Other question of mine:I
<@pboy:fedora.im>
17:37:19
I think i know what we aim at with the server part. But what are we aiming at with the client part?
<@jwhimpel:fedora.im>
17:39:10
I think we would want a user (or admin) to locally mount remote files ans use them as if they were local. It could also be used for sharing common content amongst a group of users/hosts.
<@pboy:fedora.im>
17:40:01
AH, OK. So we do kind of support to administer one's own (Fedora) workstation?
<@pboy:fedora.im>
17:40:17
I that, what about Windows and Mac workstations?
<@pboy:fedora.im>
17:40:48
If that, what about Windows and Mac workstations?
<@jwhimpel:fedora.im>
17:41:56
I would think the server part would in-scope for Mac and Windows, but I think the client part would be out-of-scope for Mac and Windows.
<@mowest:fedora.im>
17:42:38
In homelab the client NFS could be a media player system using the remote NFS share the source of the media.
<@jwhimpel:fedora.im>
17:42:58
I know ansible can run on both Mac and Windows, but it makes the role much more complicated.
<@pboy:fedora.im>
17:43:31
Yeah, OK. And I think it could b e another Fedora Server (in a VM or container), that want's to use a ressource.
<@mowest:fedora.im>
17:43:54
Peter Boy: Yes
<@jwhimpel:fedora.im>
17:45:07
It could be a fedora workstation or another fedora server that is acting as a client. WIth some effort, it could be a Debian family distribution acting as a workstation/server client.
<@pboy:fedora.im>
17:46:20
Agreed. So we should start with Fedora Server and Fedora Workstation as client. And then discuss how to expand.
<@pboy:fedora.im>
17:47:00
Fedora Server has NFS client already installed by default, but unconfigured. So it may be easy.
<@mowest:fedora.im>
17:47:17
I agree to focus just on Fedora Server and Workstation especially for NFS.
<@jwhimpel:fedora.im>
17:47:31
I think down the road, we should also consider sharing via CIFS (Samba) also.
<@pboy:fedora.im>
17:48:55
John Himpel: agreed! But there are some additional tricky issues to cope with, e.g. complicated SELinux conflicts (at least in the current default Samby installation and configuration). But let's talk about this later.
<@jwhimpel:fedora.im>
17:49:31
Agreed this should investigated later.
<@pboy:fedora.im>
17:50:23
OK, so for now we are looking forf the startup info so we can start to use and test and gather some practial experiences.
<@pboy:fedora.im>
17:50:50
I think, we should skip to topix 4 now. WE have to start release testing, too.
<@pboy:fedora.im>
17:50:57
Any objections?
<@mowest:fedora.im>
17:51:06
No objections
<@jwhimpel:fedora.im>
17:51:16
Go for it!
<@pboy:fedora.im>
17:51:21
!topic 4. Testing Release 41
<@pboy:fedora.im>
17:51:28
no tracking issue yet
<@pboy:fedora.im>
17:51:36
Release 41 has been branched so it is time to plan our testing efforts.
<@pboy:fedora.im>
17:51:36
<@pboy:fedora.im>
17:51:36
Do we want to proceed as with last release or any different ideas?
<@eseyman:fedora.im>
17:52:50
Reminder: Current Beta release date is 2024-09-17
<@pboy:fedora.im>
17:52:54
Last release: eseyman watched the change list for possible issues and we had a table who does what manual testing.
<@pboy:fedora.im>
17:53:24
Emmanuel Seyman: Yes, we really should start!
<@jwhimpel:fedora.im>
17:54:09
I can test upgrade VM F40->F41 and new install in VM of F41 (via anaconda)
<@eseyman:fedora.im>
17:54:39
Changes should have been completed yesterday. I can make another list, this time for F41
<@mowest:fedora.im>
17:54:58
I could test a standard install img on hardware
<@pboy:fedora.im>
17:55:01
Emmanuel Seyman: 1++!
<@pboy:fedora.im>
17:55:31
OK, so I take it we use the same procedure as last release.
<@pboy:fedora.im>
17:56:12
So I'll create a tracking issue and make a copy of the table we used and the we can adopt that table.
<@pboy:fedora.im>
17:56:28
Do we agree about that?
<@mowest:fedora.im>
17:56:55
Agree
<@pboy:fedora.im>
17:57:03
OK. Then
<@eseyman:fedora.im>
17:57:31
Agreed
<@pboy:fedora.im>
17:57:33
!action eseyman will monitor the change list for possible issues which may affect Server
<@pboy:fedora.im>
17:58:07
!action pboy will create a tracking issue and copy the table of F 40 as a starting point.
<@pboy:fedora.im>
17:59:56
Question: mowest Could you discuss with AdamW how we should use the result tables which are included in each test circle? Last time we discussed this, nobody of us had an idea how to use it and how to contribute
<@mowest:fedora.im>
18:00:40
I will reach out to him unless someone else is more familiar.
<@pboy:fedora.im>
18:00:55
mowest: ++1
<@zodbot:fedora.im>
18:00:56
pboy gave a cookie to mowest. They now have 7 cookies, 3 of which were obtained in the Fedora 40 release cycle
<@pboy:fedora.im>
18:01:45
!action mowest will reach out to AdamW and discuss how we can use the result tables.
<@pboy:fedora.im>
18:02:27
OK, we are running out of time. So for a short opportunity:
<@pboy:fedora.im>
18:02:36
!topic 5. Open Floor
<@pboy:fedora.im>
18:02:54
Anything here to share or to ask?
<@mowest:fedora.im>
18:03:55
None from me
<@eseyman:fedora.im>
18:04:08
Nothing here
<@pboy:fedora.im>
18:04:23
OK, obviously nothing to pick up here.
<@pboy:fedora.im>
18:04:35
!endmeeting