Den fre 24 nov. 2023 kl 06:15 skrev Miroslav Svoboda <miroslav.svoboda@xxxxxxxxx>: > please, is it better to reduce the default object size from 4MB to some > smaller value for the rbd image where there will be a lot of small mail > and webhosting files? Many of the defaults for rbd caching seems to be way above the 4M mark on this page: https://docs.ceph.com/en/mimic/rbd/rbd-config-ref/ so I think that most of the small writes would be combined into neat larger writes against the OSDs whenever flush needs to happen. I would make sure the fs-cache gets to do its job for the machine that does the RBD mounts, and then have the rbd caches help out so that you don't get a 4M write for each small file written. So in short, my guess is "no, don't change how it gets stored on disk" but rather throw some RAM at the problem instead. -- May the most significant bit of your life be positive. _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx
 |