Am 4/7/25 um 10:16 schrieb Massimo Sgaravatto:
Did I understand correctly? But doesn't this mean to have a short period of time when the cluster is in error state (in 2, during the restart) and therefore with possible problems on the cephfs clients ?
Your understanding of the procedure and its consequences is correct. The CephFS clients will block IO for the time the remaining MDS is restarted. This is similar to a restarting NFS server. An update to the MDS may affect the metadata pool, therefor it has to be done this way. Regards -- Robert Sander Linux Consultant Heinlein Consulting GmbH Schwedter Str. 8/9b, 10119 Berlin https://www.heinlein-support.de Tel: +49 30 405051 - 0 Fax: +49 30 405051 - 19 Amtsgericht Berlin-Charlottenburg - HRB 220009 B Geschäftsführer: Peer Heinlein - Sitz: Berlin