Currently, it'd be quite easy to break things without realising.
If a bad enough config change is made, it'll prevent the YAML config from being loaded, preventing tickets from being created. Unless kube jobs are being monitored (and they should be, really) it may go unnoticed for quite some time.
It'd be good to provide some kind of validation mode which can be invoked to check that a config change is OK.
Activity
08-Sep-24 16:05
assigned to @btasker
10-Sep-24 18:16
mentioned in commit 91e5f5fc6d093ce0b7c00e1c8394a65f9312deb4
Message
feat: implement dry-run mode utilities/gitlab_recurring_issue#9
10-Sep-24 18:16
Decided to go all out and create a dry-run mode.
When the environment variable
DRY_RUN
is set totrue
the system won't create tickets, it'll just print details of the ticket it would have created:10-Sep-24 18:17
However, this does mean that dry run will only be useful for checking stuff that would have been scheduled for today.
We should also add a flag that can be used to force time constraints to pass.
That way, it'd be possible to do something like
to validate the output of everything.
10-Sep-24 18:19
mentioned in commit e94896045d72f6648e90736dfdb78af62745dcee
Message
fix: add force-run mode (utilities/gitlab_recurring_issue#9)
This allows schedules to be ignored. It can be used with or without DRY_RUN mode