What Will Happen If Role Movement / Failover Failed During CAU?
Good day!
There is a charming feature introduced in Windows Failover Cluster 2012, that is something called Cluster Aware Updating (CAU).
It is designed to provide a reliable and automated method of patching clusters.
Althogh it is currently not in the CAU Frequently Asked Question list: https://technet.microsoft.com/en-us/library/hh831367.aspx.
As CAU becomes more and more popular in real customer production environments, some customers may consider the following CAU disaster recovery scenario:
What Will Happen If Role Movement / Failover Failed During CAU?
As per CAU design:
There is a -MaxRetriesPerNode parameter (default: 3) which will retry failures on a particular node.
If a node’s retries are exhausted, CAU considers the node to be “failed”, and moves on to the next node.
Once the -MaxFailedNodes limit is hit, the run will be stopped.
(The -MaxFailedNodes defaults to something like (N/3)-1, rounded down.)
Both -MaxRetriesPerNode and -MaxFailedNodes could be customized based on customer actual requests.
To customize these threshold, we can:
1. Use the CAU powershell cmdlets:
https://technet.microsoft.com/en-us/library/hh847223(v=wps.630).aspx
EXAMPLE:
This example performs a scan and a full Updating Run on the cluster named CONTOSO-FC1. This cmdlet uses the Microsoft.WindowsUpdatePlugin plug-in and requires that all cluster nodes be online before the running this cmdlet. In addition, this cmdlet allows no more than three retries per node before marking the node as failed, and allows no more than one node to fail before marking the entire Updating Run as failed. This cmdlet is performed without confirmation prompts.
Windows PowerShell
PS C:\> Invoke-CauRun -ClusterName CONTOSO-FC1 -CauPluginName Microsoft.WindowsUpdatePlugin -MaxFailedNodes -1 -MaxRetriesPerNode 3 -RequireAllNodesOnline -Force
2. Use the "Configure Self-Updating Options Wizard"
Hope this can make sense.
Best Regards,
Chengwei