Have been revisting using ceph after trialing it several years back (ok, but a headache to manage and performance sucked on the limited hardware we had).

Wow, you've really put a lot of effort integrating it into proxmox, that UI makes the setup and monitoring so easy. Outstanding work. And the Nautilus features add two key things I really like about zfs - transparent compression and checksumming. Bluetore does seem to have much better performance.

Seems pretty solid to, due to my sleep deprived state, I managed to crash/hard reboot the entire cluster *twice* today, but ceph recovered flawlessly with no loss both times, and HA brought up my critical VM's with no intervention (pfSense router, AD and SQL Server).



pve-user mailing list

Reply via email to