Here are tasks to perform when killing Taskotron: [✓] Send out final EOL info (I prepared a blogpost) [✓] Disable and stop fedmsg-hub, buildmaster, buildslave services, taskotron web frontend, cron jobs. [✓] Back up Taskotron-related Ansible files from Fedora Infra (either their full snapshot in our qa-ansible project, or pick just Taskotron files and store them somewhere). [✓] Mark all project repos as obsolete/archived [✓] Update Taskotron-related wiki pages [✓] Remove Taskotron-related docs from Fedora Infra docs [✓] Redirect https://taskotron.fedoraproject.org to https://fedoraproject.org/wiki/Taskotron and include a link to resultsdb frontend, in case look for it [✓] Find a new owner for ResultsDB deployment - somebody else's problem now [✓] Kill Taskotron boxes (VMs and bare metal) [✓] Remove Taskotron-related Ansible files from Fedora Infra [✓] Retire Fedora packages (This one I'm not completely sure about, the library functions might still be useful to someone? Perhaps even our own projects? Or it could make it easier to port existing tasks to Fedora CI in glorious future? But in that case, it might be good to strip down the packages to the minimum necessary functionality and perhaps rename them. So we might fork e.g. libtaskotron to libfedoraqa and put all useful modules in there.)
If you know of something else, please add it, I'll work on it.
I sent out the final EOL info, see #279.
I've created a taskotron-infra project and copied over hopefully all Taskotron-related Ansible files from Fedora Infra: https://pagure.io/taskotron/taskotron-infra
If something is missing, shout. This list of files will be handy later when we want to remove all Taskotron files from Fedora Infra.
I marked all Taskotron-related pagure projects as archived, except for testcloud, resultsdb* and task-abicheck (in which I created a ticket).
All non-resultsdb services have been stopped. I'm waiting until monday to delete VMs since it isn't urgent.
I updated hopefully all Taskotron-related wiki pages. @tflink, you can now create the redirect, thanks.
I created a PR for Infra docs here: https://pagure.io/infra-docs/pull-request/180 (I archived the SOP in https://pagure.io/taskotron/taskotron-infra).
@tflink, please review (and commit, if looks fine), thank you.
I created a PR for Infra ansible here: https://pagure.io/fedora-infra/ansible/pull-request/55
@tflink, please review (and commit, if it looks fine), thanks.
to be clear, since notifications went out: I put my last comment in the wrong window and wasn't meant for this issue
@tflink Hey, did you have time to look at the pending tasks? If you're too busy, do you have some idea when you might have time for it? Especially with the redirect, it makes sense to set it up soon, while people still might look for it. If we take a month to set it up, people will learn not to visit that address in the meantime.
I've been working on retiring machines and getting redirects into place.
The VMs are all retired, I haven't started on the bare metal boxes. I tried putting redirects into place today and while I didn't break anything, the redirects do not seem to be working so I'll have to take another shot at it :-(
Just to create a readable summary, the pending PRs are: https://pagure.io/infra-docs/pull-request/180 https://pagure.io/fedora-infra/ansible/pull-request/55
Redirects are in place. Something must have gone wrong when I tried to change them yesterday - I didn't have to change anything else in infra's ansible to get them working.
Still working on the HW, will get to the PRs once the metal machines are decomissioned
All solely-taskotron related VMs and bare metal machines are dead and all taskotron roles (excluding resultsdb) have been removed from infra's ansible repo.
I'm unclear on exactly who is going to own resultsdb going forward but I think it's CPE, I'll ask for clarification to make sure I haven't misunderstood something.
I still haven't gone through the docs repo, will be doing that later this week. Once that's done, I think we can probably close this ticket.
Can you check whether there's anything present in this PR that you forgot about, or whether that PR can be cancelled? https://pagure.io/fedora-infra/ansible/pull-request/55
I still haven't gone through the docs repo, will be doing that later this week.
It has been merged already: https://pagure.io/infra-docs/pull-request/180 A review is still welcome.
All the PRs were merged.
@tflink, is there a new owner for the ResultsDB deployments?
I think so, but not 100% certain. I'll check to see how that has ended up
I spoke with @pingou on IRC and while CPE has been handling the Fedora resultsdb deployments, a final decision on who's going to own them long term has yet to be made
The ownership question is being tracked in an infra ticket
Thanks, I'll consider that as 'done' from our side and mark it as completed.
I asked about people's thoughts on retiring Taskotron-related Fedora packages and there was an agreement that we should retire everything that is possible: https://lists.fedoraproject.org/archives/list/qa-devel@lists.fedoraproject.org/thread/JJVQ6UV3IAR75OO25ZSINFFAL6M5SI3X/
I'll work on that.
libtaskotron, taskotron-trigger and execdb have been retired. python-mongoquery has been orphaned (including an announcement).
libtaskotron
taskotron-trigger
execdb
python-mongoquery
And I believe this concludes this ticket. It's time to say the final goodbye to Taskotron.
<img alt="robot-heaven.png" src="/taskotron/issue/raw/files/1abaf6cbf2821f06b22256de2f37da61f299f13b4c35018c6130aac9a762dfd6-robot-heaven.png" />
Metadata Update from @kparal: - Issue close_status updated to: Fixed - Issue status updated to: Closed (was: Open)
Wait, does that mean that taskotron actually went to robot hell? <img alt="robotdevil.jpg" src="/taskotron/issue/raw/files/4a65b4bf232668f3aac1515b08a00d8e8de73574c2a1eb347dab00255d43b1d1-robotdevil.jpg" />
Login to comment on this ticket.