[10:54:48] !log lucaswerkmeister@tools-bastion-13 tools.cdnjs regenerated GitHub token (cdnjs-index/tokenfile) [10:54:50] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.cdnjs/SAL [12:04:05] !log lucaswerkmeister@tools-bastion-13 tools.cdnjs kubectl delete pod update-index-28687937-tqlqw # was apparently stuck for 16 days [12:04:07] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.cdnjs/SAL [12:05:07] !log lucaswerkmeister@tools-bastion-13 tools.cdnjs rotated + compressed update-index.err again [12:05:08] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.cdnjs/SAL [12:12:36] toolforge jobs question: can I manually start a run of an existing scheduled job? [12:12:54] (i.e. the equivalent of `systemctl start something.service` where `something.timer` would be the schedule) [12:13:30] if I try to `toolforge jobs run` it manually it complains that a job under that name exists already; but if I change the name then the job will potentially run twice in parallel if the manual run doesn’t finish before the schedule triggers the next time [12:17:40] !log lucaswerkmeister@tools-bastion-13 tools.cdnjs kubectl create job --from=cronjob/update-index update-index # manual run to see if it works better now [12:17:42] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.cdnjs/SAL [12:18:08] ^ found this via https://stackoverflow.com/a/50041304/1420237 – it probably still has the “run twice in parallel” problem but I’ve got a few hours before the schedule triggers again, so let’s just see what happens :) [13:14:17] @lucaswerkmeister: `toolforge jobs restart` [13:16:07] ah, thanks! [15:09:11] !log lucaswerkmeister@tools-bastion-13 tools.cdnjs FTR, the manual update-index job completed 23m ago, so about 2½ hours of runtime apparently [15:09:13] Logged the message at https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools.cdnjs/SAL