Hi Victor, that's true for Ceph releases prior to Octopus. The latter has some improvements in this area.. There is pending backport PR to fix that in Nautilus as well: https://github.com/ceph/ceph/pull/33889 AFAIR this topic has been discussed in this mailing list multiple times. Thanks, Igor On 3/27/2020 10:56 PM, victorhooi@yahoo.com wrote:

Hi,

I'm using Intel Optane disks to provide WAL/DB capacity for my Ceph cluster (which is part of Proxmox - for VM hosting).

I've read that WAL/DB partitions only use either 3GB, or 30GB, or 300GB - due to the way that RocksDB works.

Is this true?

My current partition for WAL/DB is 145 GB - does this mean that 115Gb of that will be permanently wasted?

Is this behaviour documented somewhere, or is there some background, so I can understand a bit more about how it works?

Thanks, Victor _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io