Sorry about the noise with the alerts. After deploying, I discovered a
few small errors in the timer/service definitions that have been
fixed.
Finally, there is an error in Wikitech that is causing `mwscript
runJobs.php` to fail frequently (T218935). That was being hidden in
the noise with cronjobs as was surfaced by adopting timers, so I think
that's a win. That job has been downtimed in Icinga for now.
On Thu, Mar 21, 2019 at 11:00 AM Giovanni Tirloni
<gtirloni(a)wikimedia.org> wrote:
Hi,
Just a heads up, I've merged a change that converts OpenStack admin cronjobs to
systemd timers.
List all timers:
```
root@cloudcontrol1003:~# systemctl list-timers --no-pager
NEXT LEFT LAST PASSED UNIT
ACTIVATES
Thu 2019-03-21 14:00:00 UTC 6min left Thu 2019-03-21 13:50:01 UTC 3min 31s ago
designate_floating_ip_ptr_records_updater.timer
designate_floating_ip_ptr_records_updater.service
Thu 2019-03-21 14:15:00 UTC 21min left n/a n/a
glance_rsync_images.timer glance_rsync_images.service
Thu 2019-03-21 14:20:00 UTC 26min left n/a n/a
keystone_delete_expired_tokens.timer keystone_delete_expired_tokens.service
Thu 2019-03-21 14:30:00 UTC 36min left n/a n/a
keystone_novaobserver_delete_tokens.timer
keystone_novaobserver_delete_tokens.service
Thu 2019-03-21 14:40:00 UTC 46min left n/a n/a
keystone_novaadmin_delete_tokens.timer keystone_novaadmin_delete_tokens.service
Fri 2019-03-22 13:30:52 UTC 23h left Thu 2019-03-21 13:30:52 UTC 22min ago
systemd-tmpfiles-clean.timer systemd-tmpfiles-clean.service
6 timers listed.
```
Now all logs can be accessed by `journalctl -u $unit` and Icinga can monitor failures
as well.
If you find any issues, please let me know.
Regards,
--
Giovanni Tirloni
Operations Engineer
Wikimedia Cloud Services
--
Giovanni Tirloni
Operations Engineer
Wikimedia Foundation