Description of problem:
separating the inactive status for storage domain to inactive and maintenance,
maintenance - the user moved the domain to maintenance
inactive - the domain is inactive due to a problem
for more details: BZ 592745
If I could just understand why maintenance should be different than detach.
We have too many states, which the user does not even understand the difference between.
detached - the domain is not part of any storage pool / DC
inactive - the domain is not accessible (failure of some kind), if it is sorted out, it should be automatically activated without user intervention
maintenance - the domain should not be accessed even if it is accessible (user stated that the domain should not be used right now) - but it cannot be attached to a different DC.
active - everything is fine
If you want to reduce the number of states then it's fine by me to unite detach and maintenance.
This RFE is influencing the BE ability for automatic recovery.
In the current status when a storage domain is Inactive there is no indication if it is as result of an error or if the user put it to maintenance and the domain should not be accessed.
I would give this a high priority because we encountered many issues that could not be fixed because of this.
Jenkins has found a scenario where this patch isn't working:
Deactivating an ExportDomain turn to maintenance but then to inactive, just like the old design.
please check the following rest logs to follow the real deactivate process:
After running some manual tests on jenkins, i found out that this bug reproduce only on Export domain and ISO domain
Created attachment 557671 [details]
How is the bug in MODIFIED state without a link to a patch?
sorry, I forgot to add them I guess.
closing ON_QA bugs as oVirt 3.1 was released: