A cluster won't trigger a failover unless there's an actual issue with one of the cluster's components (software or hardware). It will perform a basic recovery step, and the affected resource will fail over to another node because of the following possible causes:
-
Resource failure
-
Networking issue such as node eviction
-
Cluster Shared Volume (CSV) disk failure
Troubleshooting checklist
-
Identify the occurrence timestamp in the System Event Log. Then, search for events about the source Microsoft-Windows-FailoverClustering and check for Event ID 1069, 1146, or 1230.
-
Match the time zone of the System event log to the GMT time zone in the cluster log.
[!NOTE] To quickly find the time zone difference, search for
The current time is. -
Navigate to the occurrence timestamp in the cluster log and identify the corresponding line. You may find an error such as:
-
Resource <name> IsAlive has indicated failure -
IsAlive sanity check failed
[!NOTE] The error can be different depending on the issue.
-
-
Scroll up in the cluster log and try to identify if there's any other error that might be the actual cause.
-
Scroll down in the cluster log and search for
Group moveorMove of groupfor the affected resource. Take note of the exact timestamp and the destination node. -
Switch over to the cluster log of the destination's node and check the resource's behavior when it's online. If the resource manages to come online, you'll find the following log:
-
Resource <name> has come online -
Group move for <name> has completed
Otherwise, you'll find the following log:
Online for resource <name> failed -
More information
For more information, see the following articles:
Source: Microsoft Learn / IT Pro troubleshooting by Microsoft and contributors (upstream author metadata: kaushika-msft). Documentation licensed CC BY 4.0. Code samples are covered by the MIT License. Converted from Markdown, with links normalized and embedded media omitted.
