Test Scenarios for Backups & Restore

Test Setup

• create vol bak and attach to node
• connect to node via ssh and issue dd if=/dev/urandom of=/dev/longhorn/bak status=progress
• keep the dd running while doing all the tests below, that way you constantly have new data when backing up
• setup recurring backups every minute with retain count of 3
• do all the tests for each currently supported backup store nfs, s3

#1341 concurrent backup test

• Take a manual backup of the volume bak while a recurring backup is running
• verify that backup got created
• verify that backup sticks around even when recurring backups are cleaned up

#1326 concurrent backup creation & deletion

This one is a special case, were the volume only contains 1 backup, which the user requests to delete while the user has another backup in progress. Previously the in progress backup would only be written to disk after it's completed while the delete request would trigger the GC which then detects that there is no backups left on the volume which would trigger the volume deletion.

• create vol dak and attach to the same node vol bak is attached
• connect to node via ssh and issue dd if=/dev/urandom of=/dev/longhorn/dak status=progress
• wait for a bunch of data to be written (1GB)
• take a backup(1)
• wait for a bunch of data to be written (1GB)
• take a backup(2)
• immediately request deletion of backup(1)
• verify that backup(2) completes succesfully
• verify that backup(1) has been deleted
• verify that all blocks mentioned in the backup(2).cfg file are present in the blocks directory.