Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
This article provides an overview of different Azure Local update settings that control how updates are applied. The default settings provide a balanced experience of update run time versus potential workload impact. Use the available controls to fine tune and optimize the update behavior to your servicing requirements.
Prioritization
When updating Azure Local, the process live migrates virtual machines between cluster member machines to ensure workloads don't experience any downtime. Live migration failures are rare but can happen, for example, due to VM misconfigurations like attaching a virtual image (ISO) stored on a local storage path instead of a shared location (CSV).
The default solution setting prioritizes completing the update over workload availability when live migrations fail.
By using the following steps, you can change this behavior so that when live migrations fail, the update process aborts.
# Prioritize workload uptime over update completion
Enable-UpdateSetting -Name SkipForceDrain
To confirm if the setting is enabled, run this command:
# Confirm setting status
Get-UpdateSetting -Name SkipForceDrain
To switch back to the default behavior, run the following command:
# Rollback to default behavior
Disable-UpdateSetting -Name SkipForceDrain
To confirm the feature is disabled, run the following command:
# Confirm setting status
Get-UpdateSetting -Name SkipForceDrain
VM failback
Azure Local update enables maintenance mode for each node to ensure no VMs run on it before patching. After patching, the process moves the VMs running on the node back to their original location before disabling maintenance mode. To optimize the total update duration, Azure Local update disables this setting for clusters larger than 16 nodes.
For clusters up to 16 nodes, you can disable the VM failback to reduce the total update time.
To change this behavior so that VMs no longer fail back before disabling maintenance mode, use the following steps.
# Prioritize workload uptime over update completion
Disable-UpdateSetting -Name VMFailback
To enable the VM failback, run the following command:
# Prioritize workload uptime over update completion
Enable-UpdateSetting -Name VMFailback
Note
- These commands don't work on clusters with more than 16 nodes.
Next steps
Learn more about Understanding update phases
Review Troubleshooting updates