SwitchIfDead_Alert seen after removing nodes from cluster
Applies to
- ONTAP 9
- Switched clusters
- BES-53248 cluster switches
Issue
- Switch-Health shows "degraded"
::*> system health subsystem show Subsystem Health ----------------- ------------------ SAS-connect ok Environment ok Memory ok Service-Processor ok Switch-Health degraded CIFS-NDO ok Motherboard ok IO ok MetroCluster ok MetroCluster_Node ok FHM-Switch ok FHM-Bridge ok SAS-connect_Cluster ok
- Running "system health alert show" shows SwitchIfDead_Alert and is seen for non-ISL port(s):
Node::> system health alert show
Node: Node-01
Alert ID: SwitchIfDead_Alert
Resource: Ethernet1/7
Severity: Major
Indication Time: Tue Mar 25 17:19:34 2025
Suppress: false
Acknowledge: false
Probable Cause: The interface
Cluster-sw2(FOCXXXXXXXX)/Ethernet1/7 is up,
but it is not passing unicast traffic.
Possible Effect: The lack of transmit or receive traffic reported by
the switch can indicate issues with port-level or
switch-level data packet forwarding.
Corrective Actions: 1) Check interface counters and neighbor discovery information to verify the port's activity.
2) Check cluster and switch health status.- Event logs report the following alert:
[Node-01: cshmd: hm.alert.raised:alert]: Alert Id = SwitchIfDead_Alert , Alerting Resource = Cluster-switch(XXXXXXXXXX)/Ethernet1/7 raised by monitor ethernet-switch- The reported ports were previously connected to the cluster ports of some nodes.
- Those nodes have been recently removed from the cluster, but are still cabled to cluster switches.
- Switch ports used by old nodes are still admin up and healthy.