tested xtrabackup 2.3.4 and am seeing the same problem.
N.B in testing I have removed the upgrade play to just focus on the initial cluster build issues. Also, I have set the number of retries on the mysql restart handler to 1, and removed any secondary handlers from being called. There is a different, separate, issue that you will see on subsequent attempts if the first one fails, and so I did this to isolate the issue to the initial failure. Changes I made can be found in this fork:
tested xtrabackup 2.3.4 and am seeing the same problem.
N.B in testing I have removed the upgrade play to just focus on the initial cluster build issues. Also, I have set the number of retries on the mysql restart handler to 1, and removed any secondary handlers from being called. There is a different, separate, issue that you will see on subsequent attempts if the first one fails, and so I did this to isolate the issue to the initial failure. Changes I made can be found in this fork:
https:/ /github. com/mancdaz/ openstack- ansible- galera_ server/ tree/1569446