Enhancing Cluster Node Suspension when multiple UPS are managing Hyper-V failover cluster nodes.

Issue:

Whenever PCNS is managing Hyper-V failover cluster nodes using separate UPSs, there is a chance that when suspending, improper handling can lead to:

- Disks are not failing over correctly

- Virtual Machines (VMs) are becoming inaccessible

- The second node enters a paused state if the first node fails

Product:

PowerChute Network Shutdown version 5.2

Environment:

PowerChute Network Shutdown on Microsoft Hyper-V.

Cause:

Change required to PowerChute bundled script.

Solution:

This article outlines the usage of the attached script for the reliability of VM failover and accessibility during node suspension events. The attached script enhances the reliability of VM migration and cluster node suspension, significantly improving failover behaviour and system stability.

To mitigate these issues, please use the attached script, which enhances the reliability of VM migration and cluster node suspension, and it significantly improves failover behaviour and system stability.

To do this, replace “C:\Program Files\APC\PowerChute\group1\psScripts\suspend-host.ps1” with attached signed PowerShell script. Make a copy of the existing suspend-host.ps1 script before overwriting it.

Limitation

- VM migration will occur during the draining of the first node, not based on VM Priority groups.

- This means VMs cannot be prioritized for migration during the suspension process.

Released for: Schneider Electric Canada