It is possible that the resource order required a continuous stratch of the tape. Changing the resource order might have backed up the first resource first and the second resource ran out of tape space.
The job must have waited for a long time for you to put in another tape and eventually the job engine service must have failed.
As long as the backup is working it is great. However if you want to backup server2 first. Check the tape capacity and media set(overwrite protection and append protection)settings. For example start the job in a scratch media.
Hope this helps.
