Testing glusterfs again :) and trying to improve performance. I setup dm-cache with a 60GB sdd partition on both nodes, write-back enabled.
Testing performance inside a VM (only vm on node) using Crystal Disk Mark - sustained and random reads got a huge boosts, but writes were only improved by a few percent, which was surprising. Might it be because the VM image is to large for the cache? -- Lindsay
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ pve-user mailing list [email protected] http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user
