[Wikidata-bugs] [Maniphest] [Commented On] T235069: Termbox in labs should be able to recover crashes

gerritbot added a comment. Change 559967 **merged** by jenkins-bot: [wikibase/termbox@master] Make restarter.sh more robust to docker crashes https://gerrit.wikimedia.org/r/559967 TASK DETAIL https://phabricator.wikimedia.org/T235069 EMAIL PREFERENCES

[Wikidata-bugs] [Maniphest] [Commented On] T235069: Termbox in labs should be able to recover crashes

gerritbot added a comment. Change 559967 had a related patch set uploaded (by Ladsgroup; owner: Ladsgroup): [wikibase/termbox@master] Make restarter.sh more robust to docker crashes https://gerrit.wikimedia.org/r/559967 TASK DETAIL https://phabricator.wikimedia.org/T235069 EMAIL

[Wikidata-bugs] [Maniphest] [Commented On] T235069: Termbox in labs should be able to recover crashes

Ladsgroup added a comment. Why nothing is as simple as it should be :/ In order to reproduce the issue, I hard restarted wikidata-misc but "sudo service docker restart" doesn't fix the problem while `sudo docker stop 157bb9c9303f` does. That's interesting. TASK DETAIL

[Wikidata-bugs] [Maniphest] [Commented On] T235069: Termbox in labs should be able to recover crashes

Ladsgroup added a comment. because reload and restart are different things in systemd: https://askubuntu.com/a/479374 TASK DETAIL https://phabricator.wikimedia.org/T235069 EMAIL PREFERENCES https://phabricator.wikimedia.org/settings/panel/emailpreferences/ To: Ladsgroup Cc: Addshore,

[Wikidata-bugs] [Maniphest] [Commented On] T235069: Termbox in labs should be able to recover crashes

Ladsgroup added a comment. In T235069#5727949 , @Addshore wrote: > Did a quick bit of googling in the campsite story time, perhaps this is what is needed? > >> PropagatesReloadTo=, ReloadPropagatedFrom=¶ >> A space-separated list

[Wikidata-bugs] [Maniphest] [Commented On] T235069: Termbox in labs should be able to recover crashes

Stashbot added a comment. Mentioned in SAL (#wikimedia-cloud) [2019-12-18T19:47:16Z] hard reboot of wikidata-misc to test recovering from crashes T235069 TASK DETAIL https://phabricator.wikimedia.org/T235069 EMAIL PREFERENCES

[Wikidata-bugs] [Maniphest] [Commented On] T235069: Termbox in labs should be able to recover crashes

Addshore added a comment. Did a quick bit of googling in the campsite story time, perhaps this is what is needed? > PropagatesReloadTo=, ReloadPropagatedFrom=¶ > A space-separated list of one or more units where reload requests on this unit will be propagated to, or reload requests on

[Wikidata-bugs] [Maniphest] [Commented On] T235069: Termbox in labs should be able to recover crashes

Tarrow added a comment. I *think* that the problem was that the docker demon had crashed not that the termbox container had crashed. I think this was probably because the machine was rebooted at some point. I'm not sure we want the updater to constantly restart the docker service on each

[Wikidata-bugs] [Maniphest] [Commented On] T235069: Termbox in labs should be able to recover crashes

Ladsgroup added a comment. In T235069#5559609 , @Pablo-WMDE wrote: > Surprised (after a full 3 seconds of thinking about it) this is not covered by this

[Wikidata-bugs] [Maniphest] [Commented On] T235069: Termbox in labs should be able to recover crashes

Pablo-WMDE added a comment. Surprised (after a full 3 seconds of thinking about it) this is not covered by this . TASK DETAIL