Currently, we have a hard-coded list of pod events that will result in a workspace failing if they are encountered. However, there is no way to add additional events that should result in a workspace failing.
We could potentially add a new DWOC field UnrecoverableEvents that includes a list of pod events that should result workspace failure. It'd be nice to also include the threshold of how many times this event should occur before workspace failure, but we've had issues with event count in the past on OpenShift, and I'm not sure if that's still the case.
Additional context
We want to remove the FailedScheduling event from being in the default, hard-coded list of events that cause workspace failures if encountered. However, users need to be able to add this event back to the list of unrecoverable events.
Description
Currently, we have a hard-coded list of pod events that will result in a workspace failing if they are encountered. However, there is no way to add additional events that should result in a workspace failing.
We could potentially add a new DWOC field
UnrecoverableEventsthat includes a list of pod events that should result workspace failure. It'd be nice to also include the threshold of how many times this event should occur before workspace failure, but we've had issues with event count in the past on OpenShift, and I'm not sure if that's still the case.Additional context
We want to remove the FailedScheduling event from being in the default, hard-coded list of events that cause workspace failures if encountered. However, users need to be able to add this event back to the list of unrecoverable events.