Designs
- swh/devel/swh-scheduler #2345
Activity
marked this issue as related to swh/devel/swh-scheduler#2345
Following actions in order:
-
on scheduler node (saatchi), stop scheduler services (journal client, next-gen runners...) using the table migrated (origin_visit_stats)
-
on db node (belvedere):
- running migration scripts 29 and 30 [1]
- give guest access to new tables [2]
-
on scheduler node (saatchi):
- upgrade packages version
- restart scheduler services [3]
- including the next-gen scheduler runner not yet puppetized [4]
-
grafana dashboard updated with "deployment" tag
-
[1] 30.sql may run a long time as it's altering and updating the overall data table
-
[2] 29.sql introduces one so
grant select on all tables in schema public to guest; -
[3]
systemctl stop swh-scheduler-journal-client apt dist-upgrade -y systemctl restart gunicorn-swh-scheduler \ swh-scheduler-listener.service \ swh-scheduler-runner.service \ swh-scheduler-runner-priority.service \ swh-scheduler-journal-client.service- [4] next-gen scheduler runner are running in root tmux sessions in saatchi and worker17 for their respective usage
-
-
including the next-gen scheduler runner not yet puppetized [4]
All got done except this part ^. This needs first the following:
- D5809 to be rebased on latest master branch (v0.17)
- the saatchi venv (in swhscheduler home) to be updated with it
added state:wip label
-
including the next-gen scheduler runner not yet puppetized [4]
All got done except this part ^. This needs first the following:
- D5809 to be rebased on latest master branch (v0.17)
- the saatchi venv (in swhscheduler home) to be updated with it
D5089 commandeered, rebased, updated, pushed to saatchi in ~swhscheduler/swh-scheduler. ~swhscheduler/ve (venv updated). Scheduler runner next-gen rescheduled for npm, pypi, and git.
The same goes for D5818 (with worker17, which schedules git, hg and svn origins).
Those 2 are running on tmux sessions under root users.
removed state:wip label