ONTAP 9 FabricPool object.store.unavailable due to no available connections to capacity tier resulting in application failures
Applies to
- ONTAP 9
- FabricPool
- Retrieving data from capacity tier
- CIFS
Issue
- object.store.unavailable:EMERGENCY messages in ONTAP event logs
- Intermittent application failures on larger read workloads retrieving data from FabricPool capacity tier
- For example, restoring data from the capacity tier
- CIFS returning STATUS_INTERNAL_ERROR
- Applications like Jenkins may report "An internal error has occurred"
- Large amount of DELETE calls and some GET calls returning "OSC error: 6" in sktrace.log
- object_store_client_op statistics in ONTAP will show Connection Unavailable errors increasing during the problem time
- delete.Connection unavailable will likely be growing quickly
- get.Connection unavailable will likely be growing, but less quickly
- No issues detected on the object store provider or network to object store provider
- Similar problems could be seen due to issues at the object store provider or the connection to the object store
- This KB does not address issues that may arise due to these external factors