The Rook Ceph Operator should create/update the MDS deployments in case ofMDS_ALL_DOWN. When this is too dangerous in general, an option similar toskipUpgradeCheckswould help. Also, it might be worth considering reverting to the previous deployment if Pods do not start with changed parameters....
starting mds.ceph-filesystem-a at debug 2021-11-29T13:28:02.196+0000 7fb6a2520700 1 mds.ceph-filesystem-a Updating MDS map to version 4 from mon.1 debug 2021-11-29T13:28:03.114+0000 7fb6a2520700 1 mds.ceph-filesystem-a Updating MDS map to version 5 from mon.1 debug 2021-11-29T...