Both of those are showing 'tests passed' now?

If there's no results it doesn't know if they can be pushed or not, so you would either need to unpush and repush (triggering another test run) or use wavierdb-cli to 'wave' the tests.

I'm not sure why they were failed before and now passed though.

@pingou any ideas here?

On Wed, Feb 28, 2018 at 8:12 AM, Kevin Fenzi pagure@pagure.io wrote:

kevin added a new comment to an issue you are following:
``
Both of those are showing 'tests passed' now?

Yes. I did nothing. They are Tests Passed now and already pushed to testing.

If there's no results it doesn't know if they can be pushed or not, so you would either need to unpush and repush (triggering another test run) or use wavierdb-cli to 'wave' the tests.

I'm not sure why they were failed before and now passed though.

@pingou any ideas here?
``

To reply, visit the link below or just reply to this email
https://pagure.io/fedora-infrastructure/issue/6747

--
Regards,
Chenxiong Qi

So we have a cron job asking greenwave about its opinion on all pending updates, the outcome is then stored in the DB and shown in the UI.
If we happen to ask greenwave about an update before the tests for it (taskotron) could finished, we can end up with this situation where greenwave says it's missing results.
At the next cron run, taskotron will have finished running and reporting its test result, so greenwave will be able to make a decision.

We're planning (maybe already are, not sure) to migrate this to be fedmsg-based but the cron job would remain as a security measure, so this kinda of race-condition could still happen once in a while.

I think we can close this :)