Storage pool has an incompatible minimum I/O size

Hello,

one node from my 4 node cluster suddenly started to reject resource create with: storage pool has an incompatible minimum I/O size. This node already hosts 25 diskfull resources.

All my nodes provides zfs_thin type storage pools. Failing one has ashift 9 and 3 nodes ashift 12.

Is there any way to force failing node to utilize 4K sectors?

Why this ashift inconsistency wasn’t a problem before linstor-satellite service restart I did today? Cluster uses version 1.33.1 since February.