As briefly mentioned at the last IRC meeting, our wiki page is in a pretty embarrassing state.
Anyone who wants to get to know more about Fedora Server is likely to be put off. With a few changes, at least the worst can be fixed.
I propose the following changes, which we hopefully might be able to agree on in a short time at the upcoming IRC meeting.
> - - - - - - - - - - - - - <
„more background on this available here“
The founding initiative can be referred here
The Server working group meets on a weekly basis in #fedora-meeting-1.
The Server working group meets on a bi-weekly basis in #fedora-meeting.
The regularly scheduled meeting time is 4 PM ET on Wednesdays.
The regularly scheduled meeting time is 18:00 UTC on first and third Wednesdays.
Remove out of date notice
(Will be replaced by remarks at the individuell items)
SECTION Approved documents
Product Requirements Document - Approved by WG 17 January 2014 Approved by FESCo 22 January 201
Append: Currently under revision (may in red)
Goals for Server Role Installation - Approved 19 Nov 2013
MOVE to Archive, provisionally without replacement
Short-Term Goals and Long-Term Goals for Fedora Server
Polish Ideas (small ideas that would add a nice level of polish to the server product.
Both MOVE to archive provisionally without replacement
Persona Interviews - Interviews for Persona refinement
Both MOVE to archive provisionally without replacement
Server Lifecycle Proposal
Updates and Testing Proposal
Both either MOVE to archive provisionally without replacement
Or append Currently under revision
Server Roles Proposal
MOVE to archive provisionally without replacement
SECTION Meeting Minutes
MOVE completely to "Older meeting minutes"
FILL in the meetings since Dec. 2020
> - - - - - - - - - - - <
It's not a lot of work, but at least it fixes the roughest deficiencies.
I can make the changes, provided my FAS credentials allow it. But I don't want to stop anyone else from doing it either.
Of course, we have to produce articles on a reasonable scale. I can estimate that quite accurately and I think it is feasible. But what effort is required to actually see an article published on the site that we have agreed on here? What are the work processes, decision-making steps, evaluation, formatting, etc.?
This question came to mind when I tried to offer Fedora Magazine an article on Fedora Server and Cloud Images today. I was lost between dead links (404) and obviously unrelated forums or kanban pages and gave up. The effort was greater (for me) than writing the article itself (OK, a bit of an exaggeration). And then I remembered the docs project's recommendation for vi and docker containers for preview and format checking. Not low-effort.
Therefore, what's in the pipeline and how is it to be managed? Looking through various Fedora project pages, I remember a lot of them at quite an old state. That puzzles me. It indicates a relatively high maintenance effort that exceeds the available resources.
In the last IRC chat we had agreed to create a first draft proposal for a PRD update. Due to the "events of the season", the start of work was severely limited. I had made a start, but we were not yet able to work on it in the editorial group. So at the moment we only have just this version.
This first version is available at:
It’s a LibreOffice file that should be able to be opened on any Fedora system.
Maybe we can have a quick chat about this at tomorrow's IRC meeting. Otherwise we can discuss it here on the list. Alternatively, you can make changes or comments in the file and send the changed version to my email address. At the moment, these are the two easiest ways to move forward quickly.
m3xboy is working on bringing the text into hackmd.io. Then we would have another way of sharing.
In the next step, I would merge all the proposals into another version. Then perhaps already with alternative text passages that could be decided upon.
During our IRC meeting last Wednesday we had agreed that I would make a first draft proposal for contents of a possible Fedora Server box on the docs landing page (as to the already existing ones on IoT, CoreOS etc).
It is now available at https://pboy.fedorapeople.org/FedoraServerDocpageProposal_v1-0.odt
Please, comment here on the mailing list (if possible use the item numbers as reference) or modify the text directly and send it to me (for the time being, improvement is in progress).
I know, it’s probably the wrong list here, but I asked over on cloud list and couldn’t induce an answer. I hope the technical experts here can help me.
I’m currently working on a text how to use Fedora cloud images as VM’s for production on a standalone Fedora Server, intended as part of our work to enhance our documentation.
Specifically, it is about the use of --cloud-init option in virt-install version 3, which has become available in Fedora 33. It is quite a simple way, especially useful if you only want to create a few VMs once. And you can use it perfectly in scripts or Ansible playbooks.
I have already got it to work with CentOS cloud image.
Now I am looking for information on what the differences are between the Fedora Base Raw image and the Openstack Base qcow2 image (beyond the file format). What different usage rationale underlies the construction of the images.
The documentation I found is quite old (Fedora 25 here in docs, or from about 2016 over at https://fedoracloud.readthedocs.io).
Please could someone give me a hint where I can grap the info I’m seeking.
Yesterday a bi-weekly IRC meetings was held. Just for information and as an encouragement to participate in the discussion, I would like to summarise the most important topics. A short IRC summary is available at
the complete log at
* TOPIC: Fedora Server Product Requirement Document (PRD) Update*
A proposal is available for download at https://pboy.fedorapeople.org/FedoraServerPRD-UpdateProposal-V1-0.odt (LibreOffice format). There is ongoing work to make it available at hackmd.io as well.
Everyone is (strongly :-) ) invited to contribute ideas or assessments on the individual topics, here on the mailing list or directly in the document (and send it to the author at the moment).
Further discussion at the next meeting ( scheduled 3 Feb. 2021)
* TOPIC: Improving Fedora Server documentation and visibility *
The starting reference is the suggestion (Matthew) to add a "Fedora Server" box to the docs home page, complementing the existing IoT, Silverblue and CoreOS boxes.
The general assessment is that a shift from the wiki to the docs system would be positive.
What needs to be clarified is what is to be included there. Some ideas: Server documentation, howtos, ongoing projects and some items that are currently on the wiki page.
Some arguments to be weighed up:
- It is problematic that documentation becomes outdated more or less quickly.
- No documentation might be better than outdated documentation.
- Documentation should "grey itself out" over the years (or may be revoked automatically after x years if not updated)
- There is also documentation in various places that includes Fedora Server.
- Howtos would perhaps be better placed elsewhere, e.g. Fedora Magazine perhaps.
- Current and accurate How-To's do attract new users.
- If we add "roles" in addition to base server, then docs and possible Git repositories become more important.
- We should / must increase visibility.
Agreed: pboy will prepare a proposal on possible contents for the next meeting.
* TOPIC: systemd-oomd *
change proposal: https://fedoraproject.org/wiki/Changes/EnableSystemdOomd#How_to_test
- currently only Fedora Workstation and the KDE Spin comes with a userspace oom killer, and it's earlyoom
- the change will enable systemd-oomd on every Fedora solution that seems suitable
Broad assessment: should be fine on server
- if you have say a httpd and are running several applications on it, they will all be in the same cgroup
- maybe jBoss / wildfly are affected the same way
Agreed: The aspect would have to be well documented. Further testing.
* TOPIC: open floor *
The Wiki pages urgently need to be updated, at least the completely deprecated information simply needs to be deleted.
Agreed: Further discussion next meeting.
### Just in case anyone missed it: Everyone is strongly invited to contribute! ###
(Or as it was called in ancient Rome: ceterum censeo … )
Looking at the reboot meeting logs, looks like we initially meant to
reconvene on January 6, then pivoted to finding a next meeting time and
avoid selection bias... then with the holidays that never happened.
I've started a WhenIsGood, picking a timezone that should work for
people from the US West Coast all the way to Central Europe -- if this
doesn't work for you, please chime in and I'll get this edited:
These are one-hour slots, over a period of two weeks starting next
Monday (UTC); please find a time that will work for you regularly so we
can then just schedule a permanent meeting.
I've also slotted in a recurring meeting in the calendar that we can
use by default if we can't find consensus on a new meeting slot -- it's
two weeks after the initial Jan 6 proposal, at the same time (18:00
Michel Alexandre Salim
chat via email: https://delta.chat/
GPG key: 5DCE 2E7E 9C3B 1CFF D335 C1D7 8B22 9D2F 7CCC 04F2