Archived Etherpad for the session:
Wed, Nov 6
We may have to write a small adaptor, but other than that, yes.
Tue, Nov 5
Aye, Argo is what we call it, though it's actually many components.
Argo, Airflow, and CI
For RelEng and CI, https://www.mediawiki.org/wiki/User:LarsWirzenius/NewCI is long, but the most important parts are introduction and requirements, which aren't all that long.
Mon, Nov 4
In the hope it's useful, I'll mention I'm still logged in to Gerrit, since my previous log in on October 24.
Fri, Nov 1
Wed, Oct 30
I declare this train to be finished.
Fri, Oct 25
To be clear, I'm only reporting these logins as data points to aid debugging, not to complain. The login is easy enough for me on my laptop that it's not bothering me. I can stop reporting data points if it's not useful.
Today I didn't need to log in in the morning, but did have to just now.
Thu, Oct 24
There seems to have been a spike half a hour ago, nothing since.
There's almost a thousand of these now.
Deployed to all wikis. Logstash didn't explode immediately, but keeping an eye on it.
OK, thanks. Removing as blocker.
OK, removing as blocker. Thanks!
There's a lot of these for 1.35.0-wmf.2, in the past 24 hours, but none for wmf.3.
I have made a point of clicking the remember-me box every login while collecting data for this ticket
I too had to log in today. Debian buster, FF 60. No reboot.
Wed, Oct 23
I've been keeping a log for when I need to log in to Gerrit:
Tue, Oct 22
No more blockers, continuing to deploy to group0.
andrewbogott told me on IRC this is temporary problem for labtestwiki only and that we can ignore labtestwiki for this train.
Got as far as the scap sync of wikiversions.json to testwiki, but made a mistake and it went to labtestwiki instead, and not even there because a database failure. Train blocked. See subtask.
Started the branch cut. There are some possible issues with new Gerrit, but co-workers seem confident they don't affect this. We won't know if it works until we've tried it.
Wed, Oct 16
Data point: I tend to need to log in about once a day, based on memory. Have not kept a log, though. I use Firefox 60 on Debian, and only ever log in on that browser on this machine. I just checked, I am not currently logged in to Gerrit, but I know I was a few days ago.
Oct 7 2019
It's usually a bad idea to embed secrets into containers. It's better if the container can get the secret when it starts, but I'm not sure if our K8s cluster is set up for that. Or if these containers are run in K8s.
Oct 3 2019
I too like this proposal, and as Tyler said, the plans for future CI include more support for self-serving. I think it's the right way to go.
I continue to comment on this, while being ignorant about everything. If MW can read a secret from a file, I suggest this:
Oct 2 2019
I'm not familiar with MediaWiki internals at all, so please excuse me for asking what may be excessive basic questions. Is the following a sufficiently correct understanding: CPT wants CI to deploy an MW test instance to a test machine, and want to run integration or other tests against that machine. The tests will require a shared secret between the test and the software running on the test machine.
Oct 1 2019
Sep 16 2019
Sep 12 2019
Sep 6 2019
Out of curiosity, does the output of "sudo env" vary between the two boxes?
Sep 5 2019
Aug 13 2019
We've been discussing PoC evaluation criteria, and have come up with this spreadsheet: