[11:03:42] seems like wikibugs is down, can someone restart it? [11:32:11] and the bridge bot is double posting again 😅 [11:32:23] (wikibugs is back, thanks to whoever did that) [12:38:32] !log tools hard reboot tools-prometheus-6 [12:38:36] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/SAL [12:45:53] TheresNoTime: is eranbot 'enwiki' job intentionally still running in SGE? your comment seems to indicate it should have ben moved [12:46:52] !log lucaswerkmeister@tools-sgebastion-10 tools.bridgebot Double IRC messages to other bridges [12:46:55] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.bridgebot/SAL [13:12:35] Why am I having an issue running pip as https://www.mediawiki.org/wiki/Manual:Pywikibot/OAuth/Wikimedia says to run [13:12:54] I get "-bash: pip: command not found [13:12:55] " [13:13:01] "-bash: pip: command not found [13:13:01] " [13:14:12] taavi: urgh I didn't kill the already running job [13:15:19] On lunch at the moment but I'll kill it when I get back (or feel free to in the meantime) [13:15:58] PotsdamLamb: where are you running pip in? your laptop? it seems it's not installed/available [13:16:36] try pip3 [13:23:09] @RhinosF1 tried that no luck, same error;@dcaro no on toolforge trying to set up my instance for a bot. Following the instructions, I get to the first pip command and it doesn't work [13:24:44] PotsdamLamb: you can follow this on toolforge https://wikitech.wikimedia.org/wiki/Help:Toolforge/Running_Pywikibot_scripts [13:25:03] if you want to run scripts [13:26:25] @dcaro, so right at the top to get my tokens, it says to use pip [13:26:26] https://www.mediawiki.org/wiki/Manual:Pywikibot/OAuth/Wikimedia#Registering_your_bot_with_the_wiki_software [13:26:52] pip install mwoauth [13:27:59] ah, I see, that step is not needed (pip install + configure.py), you only need the part about registering the bot https://www.mediawiki.org/wiki/Manual:Pywikibot/OAuth/Wikimedia#Registering_your_bot_with_the_wiki_software [13:28:53] to get the tokens+secrets [13:29:31] ah ok. The documentation should be updated. I even took forever to realize you have to use python3 instead of python as the manual says since the bot only works with >= 3 [13:58:15] (killed the job ta/avi) [13:59:07] can `jobs.yaml` have `mem: 500Mi` (a la `--mem 500Mi`)? [14:00:01] I don't remember the exact syntax, but setting memory somehow is possible [14:07:46] https://gitlab.wikimedia.org/repos/cloud/toolforge/jobs-cli/-/blob/main/tjf_cli/cli.py?ref_type=heads suggests it is `mem` :D [15:21:07] !log anticomposite@tools-sgebastion-10 tools.stewardbots SULWatcher/manage.sh restart # SULWatchers disconnected [15:21:10] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.stewardbots/SAL [15:46:06] !log samtar@tools-sgebastion-10 tools.copypatrol webservice stop && webservice --mem 1G start [15:46:09] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.copypatrol/SAL [19:59:04] !log bd808@tools-sgebastion-11 tools.wikibugs Update to a3bef6e (T128868, T140881, T90339, T281055, T180293) [19:59:16] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.wikibugs/SAL [21:25:33] I am going to ask for ps to be available in toolforge images [21:44:02] TimStarling: In theory there should only be one process in a container. In theory [21:45:39] in theory all cows are spherical balls of uniform density [21:47:07] but that is not the world we live in [21:52:59] https://gerrit.wikimedia.org/r/plugins/gitiles/operations/docker-images/toollabs-images/+/refs/heads/master/bookworm-sssd/Dockerfile.template would probably be the place to add it. That cascades to all of the newest images. [21:53:10] yeah done already [21:53:26] https://gerrit.wikimedia.org/r/c/operations/docker-images/toollabs-images/+/1010690 [22:14:20] how do you feel about kubectl debug? I tried it but it is not allowed [22:18:25] I've never used it or thought about using it [22:19:40] it https://kubernetes.io/docs/concepts/workloads/pods/ephemeral-containers/ [22:22:09] would that be a way to keep debug utilities out of the main images (by putting them in a separate image for ephemeral containers instead)? [22:22:38] I have heard of ephemeral containers. serviceops blocked making a maintenance env for Toolhub on the idea that someday the main cluster will support ephemeral-containers/ [22:25:00] lucaswerkmeister: yes, although I think procps is pretty harmless to install in the base image, and it's fine to provide multiple ways to do things [22:26:25] ok, fair ^^ [22:27:43] I haven't actually used it, I'm just reading the manual, but I gather the ephemeral container will share a process namespace (so you can run ps) but not a root filesystem (so you can't inspect the state of temporary files) [22:29:26] there are some hints in the manual that ephemeral containers are not as convenient as kubectl exec, but allow you to harden the production image [22:29:39] your main image can be distroless, without a shell [22:32:24] to me it sounds similar to systemd-sysext, which also has “I’d like to have debug / dev tools now” as one of its use cases