I have a 12 node/brick gluster volume setup in disperse mode with a redundancy of 4. On a 1gbps LAN I am getting expected performance of ~75MB/s write to it.
On a client with the glusterfs volume mounted I can setup an ecryptfs mount on the gluster volume and have encrypted files and filenames on the gluster volume. This works.
However, cpu usage of the gluster process skyrockets and write speeds drop to a mere ~1.5MB/s on the ecryptfs mount.
I'm guessing ecryptfs is making a large volume of filesystem calls that is taxing to glusterfs but less so on standard filesystems. I have enabled verbose output on ecryptfs but that yielded no clues. And gluster logs don't show anything unusual at first glance.
Any suggestions at where I might look for further information or for someone who has solved this issue?
________ Community Meeting Calendar: Schedule - Every 2nd and 4th Tuesday at 14:30 IST / 09:00 UTC Bridge: https://meet.google.com/cpu-eiue-hvk Gluster-users mailing list Gluster-users@xxxxxxxxxxx https://lists.gluster.org/mailman/listinfo/gluster-users
