Hm… oddity, the job was created with a machinelimit of 1, yet check this out:
When i remoted into lapro0463, it was rendering a different job…
It seems the slave might have been shut down on that box, but when it restarted, its previous activity was not cleared in the db?
The slave probably wasn’t shut down cleanly, but the task should have been automatically cleaned up after being in the rendering state for 10 minutes (in the image, it has only been rendering for 8 minutes).
We’ll be tweaking this though to wait 1 minute instead of 10, since it’s not really necessary to wait that long. This way, these orphaned tasks should get cleaned up quicker. That’s actually how long the orphaned limit stubs are cleaned up anyways (which is why another slave was able to pick up this job). Since they will be cleaned up at the same time, you shouldn’t see this odd behavior in the future.
Cheers,
Cool, that explains it, i was confused how the other slave could pick it up! thx
