[Gluster-users] Solaris 10 client NFS performance very poor

Dan Bretherton d.a.bretherton at reading.ac.uk
Mon Jul 16 18:42:24 UTC 2012


Dear All,
I am having a lot of trouble accessing GlusterFS 3.2.6 volumes from 
Solaris 10 NFS clients.  Volumes can be mounted but I/O performance is 
so poor that the volumes are practically unusable.  Doing cd or ls can 
take a very long time and often hangs, with repeated errors like the 
following.

mars etc # cd /users/rle
NFS server glusterfs not responding still trying
NFS server glusterfs ok

I am using the mount options recommended in the GlusterFS 3.2 
Administration Guide.  I also tried adding mount options 
rsize=32768,wsize=32768 but it didn't seem to make any difference.  
These are fixed mounts; automount doesn't work at all with GlusterFS 
volumes on Solaris 10.  I have no performance problems with my Linux NFS 
clients.  Can anyone suggest a way to improve GlusterFS access from 
Solaris 10?

Regards,
Dan.



More information about the Gluster-users mailing list