[08:00:20] <wm-bot2>	 Cteam: welcome to today 🦄! Don’t forget to post your update in thread.
[08:00:20] <wm-bot2>	 Feel free to include:
[08:00:20] <wm-bot2>	 1. 🕫 Anything you'd like to share about your work
[08:00:20] <wm-bot2>	 2. ☏ Anything you'd like to get help with
[08:00:20] <wm-bot2>	 3. ⚠ Anything you're currently blocked on
[08:00:20] <wm-bot2>	 (this message is from a toolforge job under the admin project)
[08:48:14] <addshore>	 I have gone through a bunch of pahb tickets related to CLIs, and also making APIs easier to call and left some comments with thoughts
[08:48:14] <addshore>	 Of note this lead to me adding "curl" commands to the builds, envvars and jobs commands, so you can make arbitrary requests
[08:48:14] <addshore>	 I also make a kubernetes command that allows arbitrary authenticated requests to the k8s api too
[08:48:14] <addshore>	 eg. locally I can do.. tf k curl --tool wikicrowd /api/v1 --json
[08:48:57] <addshore>	 I'll likely start getting CI to build binaries in the coming days so that its easier for folks to give the CLI a go
[13:42:09] <dhinus>	 Done:
[13:42:11] <dhinus>	 * resolved T362868 toolforge k8s upgrade
[13:42:13] <dhinus>	 ** upgraded lima-kilo
[13:42:15] <dhinus>	 ** updated the wiki at https://wikitech.wikimedia.org/wiki/Portal:Toolforge/Ongoing_Efforts/Toolforge_Upgrade_Workgroup/Upgrades_Overview
[13:42:17] <dhinus>	 * T383723 clouddumps1001 temperature issues: attempted shut down
[13:42:19] <dhinus>	 ** created subtask T391369 to track the issues I observed
[13:42:21] <dhinus>	 * jobs-api code reviews (!153 is now merged)
[13:42:23] <dhinus>	 Working on:
[13:42:25] <dhinus>	 * more jobs-api code reviews
[13:42:27] <dhinus>	 * restarting work on some lower-priority tasks:
[13:42:29] <dhinus>	 ** T374953 replace wmcs-wikireplica-dns.py with tofu
[13:42:31] <dhinus>	 ** T385885 [toolsdb] Remove apt pinning
[13:50:55] <andrewbogott>	 Yesterday:
[13:50:55] <andrewbogott>	 * Worked more on magnum flakiness; it's working reasonably well for me in codfw1dev
[13:50:55] <andrewbogott>	 * Replaced cloudcontrol1005 with cloudcontrol1011
[13:50:55] <andrewbogott>	 Today:
[13:50:55] <andrewbogott>	 * Reviewing  hardware tasks/spreadsheets
[13:50:55] <andrewbogott>	 * Getting new test ceph drive online T390134
[21:20:19] <bd808>	 andrewbogott: random datapoint for you--I destroyed and recreated the deployment-prep Magnum cluster yesterday. Destroying was an adventure that actually needed manual bits via Horizon; Tofu kept timing out in several of the delete operations. Recreating "just worked" using the values previously coded into my Magnum template. I didn't try bumping to any newer version.
[21:22:07] <andrewbogott>	 bd808: My current theory is that those intermittent failures were because of running out of DB connections. I got some good clean runs after merging this silly change: https://gerrit.wikimedia.org/r/c/operations/puppet/+/1135451
[21:22:48] <andrewbogott>	 Also, interesting, possibly good news: the magnum devs have given up on heat and wrote an entirely new driver which is available in the next release.