* Address slow mysqld start upon unit reboot

* Handle race conditions in snap restarts and update status executions + clean up error logging when recovering

* Fix failing unit test

* Run format

* Use simpler query for cluster metadata using mysqlsh

* Fix issue with get_cluster_metadata query when async setup used

* Fix failing unittest

* Address PR feedback

* Remove redundant check when executing manual rejoin

* Bump libpatch of the mysql charm lib

* Increase number of log lines retrieved in assertion to account for error traces

* Address edge case in rollback test where snap refresh was unnecessary

* Use status log to determine successful rollback instead of debug logs