Select all the volumes using batch selection. Click the batch operation button Upgrade Engine, and choose the engine image available in the list. It’s the default engine shipped with the manager for this release.
Resume all workloads. Any volume not part of a Kubernetes workload must be attached from the Longhorn UI.
Live upgrade is available since v0.3.3, with the exception of upgrade from v0.5.0 to v0.6.0, v0.7.0 to v0.8.0, and v0.8.1 to v1.0.0.
The iSCSI frontend does not support live upgrades.
Live upgrade should only be done with healthy volumes.
Select the volume you want to upgrade.
Click Upgrade Engine in the drop down.
Select the engine image you want to upgrade to.
Normally it’s the only engine image in the list, since the UI exclude the current image from the list.
During the live upgrade, the user will see double number of the replicas temporarily. After upgrade complete, the user should see the same number of the replicas as before, and the Engine Image field of the volume should be updated.
Notice after the live upgrade, Rancher or Kubernetes would still show the old version of image for the engine, and new version for the replicas. It’s expected. The upgrade is success if you see the new version of image listed as the volume image in the Volume Detail page.
Clean up the old image
After you’ve done upgrade for all the images, select Settings/Engine Image from Longhorn UI. Now you should able to remove the non-default image.