Giveback/Takeover of a Fabric Pool aggregate fails, but fabric pool destination is pingable from home node
Applies to
- ONTAP 9
- ONTAP Select
- FabricPool
- Cloud Volumes ONTAP(CVO)
Issue
- Attempt to perform SFO giveback of a Fabric Pool aggregate fails with the below errors:
[Node-01: cf_giveback: gb.netra.ca.check.failed:error]: Giveback of aggregate 'aggr1' (uuid: XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX) failed due to Object store is not reachable on destination preventing object store access on the destination node.
[Node-01: cf_giveback: sfo.sendhome.subsystemAbort:alert]: The giveback operation of 'aggr1' was aborted by 'fabric pools'.
[Node-01: cf_giveback: sfo.giveback.aggrProcessTime:debug]: Time taken to destination check during giveback of the aggregate 'aggr1' was 44 milliseconds.
[Node-01: cf_giveback: sfo.giveback.failed:alert]: Giveback of aggregate aggr1 failed due to destination check failed.
[Node-01: cf_giveback: ha.giveback.sysAbort:debug]: Subsystem wafl took 12 msecs to abort giveback of aggregate 'aggr1'.
[Node-01: cf_giveback: ha.giveback.sysAbort:debug]: Subsystem lock_manager_NDO took 12 msecs to abort giveback of aggregate 'aggr1'.
[Node-01: cf_giveback: ha.giveback.sysAbort:debug]: Subsystem coredump took 12 msecs to abort giveback of aggregate 'aggr1'.
[Node-01: cf_giveback: ha.giveback.totalAbort:debug]: Total time taken to abort the giveback of aggregate 'aggr1' was 36 msecs.
OR
- Attempting to takeover a node fails with the below error:
[node1: sfo_arl_worker: sfo.tkAbort.ca.check.failed:error]: Planned SFO takeover of aggregate 'aggr1' (uuid: XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX) failed due to Object store is not reachable on destination.
cluster1::> storage failover show-takeover
Node Node Status Aggregate Takeover Status
---------- --------------------- -------------- -------------------------------
node1
Optimized takeover
by partner aborted.
aggr1 Previous takeover attempt was
aborted because of a precheck
failure. Failing module(s):
"fabric pools". Use the "event
log show -message-name
sfo.tkAbort*" command to get
more information, and follow
the provided corrective
actions.To abort the takeover,
you can giveback the
aggregates that were relocated
as part of the attempted
takeover using the "storage
failover giveback -ofnode
node1" command.If you want to
continue the takeover, address
the failures and then use the
"storage failover takeover
-ofnode node1" command.If you
want to continue with the takeover
without addressing the
failures you can do so by
using the "storage failover
takeover -ofnode
node1 -bypass-optimization true"
command. Warning: Setting the
"-bypass-optimization"
parameter to true might result
in a longer client outage
during planned takeover.
CFO aggregates Not attempted yet.
- During ONTAP upgrade, ANDU will be paused because of the timeout while performing giveback:
[Node-02: notifyd: callhome.andu.pausederr:alert]: params: {'subject': 'AUTOMATED NDU PAUSED ON NODE: Node-01', 'epoch': 'XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX'}
- Object stores go to an unavailable state.
- Fabric Pool destination is reachable from Intercluster LIFs on both nodes in the HA pair via ICMP.