Root Volume space full due to RAID Parity Inconsistency large files
Applies to
- ONTAP 9
- AFF/FAS
Issue
Root volumes on nodes are filling up rapidly, following symptoms and log outputs are observed:
- EMS/event log shows repeated RAID errors:
NOTICE raid.tetris.cksum.embed: Invalid checksum entry on Disk /aggr0_Node_aggr_name/plex0/rg0/0n.5P3 Shelf 0 Bay 5 [NETAPP X4019S172B15TNTE NA06] S/N [XXXX] UID [65593030:54601064:00253841:00000004:500A0981:00000003:00000000:00000000:00000000:00000000], block #XXX, during write operation.
NOTICE raid.rg.readerr.repair.cksum.error: Data recalculated from parity on Disk /Node_aggr_name/plex0/rg0/0n.5P1 Shelf 0 Bay 5 [NETAPP X4019S172B15TNTE NA06] S/N [XXXX] UID [65593030:54601064:00253841:00000004:500A0981:00000001:00000000:00000000:00000000:00000000]; block #XXX does not match the stored and computed checksums.
NOTICE raid.rg.readerr.repair.data: Fixing bad data on Disk /Node_aggr_name/plex0/rg0/0n.5P1 Shelf 0 Bay 5 [NETAPP X4019S172B15TNTE NA06] S/N [XXXX] UID [65593030:54601064:00253841:00000004:500A0981:00000001:00000000:00000000:00000000:00000000], disk block (DBN) XXXX, aggregate block (VBN) XXXX.
- Large files (15GB–19GB) named parity_inconsistencies* are present in /mroot/etc/raid:
::*> system shell -node <node_name> -command ls -la /mroot/etc/raid-rw-r--r-- 1 root wheel 17471671295 parity_inconsistencies.4-rw-r--r-- 1 root wheel 15993555755 parity_inconsistencies.0
- Root volume utilization spikes to 98%, risking system stability.
