Hi,
Is there anyway to configure Horde to automatically retry a failed step ( not job ) without erroring out?
It improves CI/CD stability so I’m curious if this option is available anywhere?
Could not find anything in the API documentation so any pointers would be greatly appreciated.
Thank you.
Kind Regards,
Abhi
Hello,
The recommendation is for the tool/process that is failing would handle retries internally itself as it would have the context of the specific error and if a retry should be done or not be done. For example if it was a transient error that can be retried vs an error like a compilation error due to invalid source code which would be a waste of resources to retry because the same error would happen again. Therefore for Horde itself to automatically retry is not recommended.
Matthew
Hi Abhi, if you have not already could you open a new case to report the cooking/engine bug. A new case will be required to get it routed to the appropriate team, and that will be the correct way to solve this issue.
Thank you
Hi, cool I will close this ticket.
Thanks
Matthew
Hi Matthew,
In my case the step thats failing is cooking because of an engine bug related to race condition in controlrigvm. However it almost always succeeds during subsequent run. So I guess there may be a retry option for a cook job if so I might look into that.
Kind Regards,
Abhi
Hi Matthew,
Its already open and tracked in bug tracker, here is the link to the post : Unreal Engine Issues and Bug Tracker (UE-317168)
Looks like its fixed in 5.7 so I’ll try it out