[Gluster-users] NfS-Ganesha - Performance after a while bad

Taste-Of-IT kontakt at taste-of-it.de
Tue Sep 28 12:47:45 UTC 2021

Hi all,

i have installed a 2 node distributed volume on Debian 10. Both Nodes have Raid6 and xfs as filesystem. NFS-Ganesha is configured on Node1. I actualle transfering lot of files, mostly bigger than 4GB from folders from both nodes to the new volume. Both nodes connected to the nfs share at node1. Actually i have only 1GBits connection between node1 and node2. Both nodes have 16GB RAM.

If i copy directories with lot of files and folders and also mostly files bigger than 4gb, the performance will slow down after a while. As an example, if i start to copy i got the most out of 1Gbits with approx 100MBs, but after a while it slows down to under 20MBs sometimes only 8MBs. I have nothing special configured in glusterfs and also not tuned the filesystem.

Any ideas?


More information about the Gluster-users mailing list