WAFL inconsistency during volume clone split with parallel re-key
Applies to
- AFF
- ONTAP <= 9.7
- NetApp Volume Encryption (NVE)
- Volume Clone Split
Issue
- Volume clone is created from a originating volume that is encrypted using NVE.
- The clone is splitted which also triggers a parallel re-key operation for NVE in ONTAP <9.7.
- The node running the clone-split + re-key panics:
Fri Jun 30 09:22:24 UTC [node-01: wafl_exempt09: sk.panic:alert]: Panic String: Unrecoverable metadata block (file 77, block 15765045879, fbn 66948, level 0, file type 1) in volume myVolume@vserver:bef12b50-1234-dead-babe-00a098f48f12. WAFL inconsistent. Contact NetApp technical support. in SK process wafl_exempt09 on release 9.5P9 (C)
- HA- or DR-Partner that takes over services due to the panic does triggers an identical panic.
- HA- or DR-Partner comes online again in Takeover/Switchover and the clone volume is offline, marked WAFL inconsistent.
- Metrocluster switchback or HA giveback is vetoed similar to:
Wed Jul 05 16:52:09 UTC [nacons-n11: vdb_apply_work22: wafl.mcc.sb.nso.veto.nvsave:error]: Switchback or negotiated switchover vetoed due to presence of NVSave files for 'Flexible Volume' (FSID 1377713130, UUID fb14c13d-dead-babe-affe-00a098f13777).