Switch alert reported for switch that was already removed
Applies to
- ONTAP 9
- Cluster Switch Health Monitor (CSHM)
- Cluster network switches
Issue
- System has been turned into a switchless cluster by removing the network switches previously in use
- After a reboot of the nodes, the system starts generating weekly health alerts for the switches that are no longer physically present
1/5/2022 06:17:59 node-04 ALERT hm.alert.raised: Alert Id = UnsupportedSwitch_Alert , Alerting Resource = switch-01(XXXXXXXXXXX) raised by monitor ethernet-switch
1/5/2022 06:17:59 node-04 ALERT hm.alert.raised: Alert Id = UnsupportedSwitch_Alert , Alerting Resource = switch-02(XXXXXXXXXXX) raised by monitor ethernet-switch
1/5/2022 06:17:59 node-04 ALERT hm.alert.raised: Alert Id = SwitchCommunityString_Alert , Alerting Resource = switch-01(XXXXXXXXXXX) raised by monitor ethernet-switch
1/5/2022 06:17:59 node-04 ALERT hm.alert.raised: Alert Id = SwitchCommunityString_Alert , Alerting Resource = switch-02(XXXXXXXXXXX) raised by monitor ethernet-switch
- The health of the switch subsystem is showing as degraded
Cluster::> system health config show
Node Monitor Subsystem Health
------------- ---------------------- --------------- ------------------
node-03 node-connect SAS-connect, CIFS-NDO, MetroCluster_Node
ok
node-03 controller Environment, Memory, Service-Processor, Motherboard, IO
ok
node-04 node-connect SAS-connect, CIFS-NDO, MetroCluster_Node
ok
node-04 system-connect SAS-connect_Cluster, MetroCluster, FHM-Switch, FHM-Bridge
ok
node-04 system - ok
node-04 controller Environment, Memory, Service-Processor, Motherboard, IO
ok
node-04 chassis Environment ok
node-04 ethernet-switch Switch-Health degraded
8 entries were displayed.
- Deleting the logged health alerts only provides a temporary solution
Cluster::> system health alert delete -node * -monitor ethernet-switch -alerting-resource * -alert-id UnsupportedSwitch_Alert
Cluster::> system health alert delete -node * -monitor ethernet-switch -alerting-resource * -alert-id SwitchCommunityString_Alert
- The switches are still visible under
CSHM-SWITCH-CONFIG.XML
in the Weekly Autosupports - The switches are no longer present on the CLI, so they cannot be removed from monitoring
Cluster::> system switch ethernet show
This table is currently empty.
Cluster::> system switch ethernet delete -device <switch name>
Error: There are no entries matching your query.