e0a/e0b port down leads to node panic
Applies to
- MetroCluster IP
- AFF A220
- IP switch replacement/reboot
Issue
During an IP switch replacement/reboot e0a/e0b ports go down, which might lead to node reboot:
[NodeA: intr: netif.linkDown:info]: Ethernet e0b: Link down, check cable.
[NodeA: intr: netif.linkDown:info]: Ethernet e0b-20: Link down, check cable.
[NodeA: intr: rlib.ifconfig.linkEvent:notice]: params: {'eventType': 'DOWN', 'ifname': 'e0b'}
[NodeA: intr: rlib.ifconfig.linkEvent:notice]: params: {'eventType': 'DOWN', 'ifname': 'e0b-20'}
PANIC: WAFL hung for aggr1. in SK process wafl_exempt01 on release 9.8P2 (C)