To complete checkout the sources, it takes about 6 minutes on a Linux NFS
client, but it takes about 40 minutes on Solaris 8 NFS client.
Why the NFS traffic is so slow between the RHEL NFS server and Solaris NFS
client? Are there any parameters I can adjust on the Solaris client?

BTW, I have installed the newest patch cluster for the Solaris machine, and
all machines are in the NIS environment. There is no other special
authentication. I have also tried to mount the NFS in TCP connection, but
there is no difference.

Any idea would be appreciated,
Ross

Try specifying the "rsize=8192" and "wsize=8192" option parameters on
the RHEL mount.  To find out more, do this:

$ man fstab

Cheers

--
I'm already an anomaly, I shall soon be an anachronism, and I have
every intention of dying an abuse!