Upgrade Conflict Handling test

New installation:

  1. Create a large cluster of many nodes (about 30 nodes)
  2. Install Longhorn master
  3. Create 100 volumes using volume template claim in statefulSet.
  4. Have the backup store configured and create some backups.
  5. Set some recurring jobs in the cluster every 1 minute.
  6. Observe the setup for 1/2 an hr. Do some operation like attaching detaching the volumes.
  7. Verify there is no error in the Longhorn manager.

Upgrading from old version:

  1. Repeat the steps from above test case with Longhorn Prior version.
  2. Upgrade Longhorn to master.
  3. Do some operation like attaching and detaching the volumes.
  4. Verify there is no error in the Longhorn manager.

Success if: install/upgrade successfully after maximum 15 minutes.

Fail if: The upgrading is blocked after about 17-20 managers are up.

[Edit]