linux<->solaris NFS performance questions

From: Ivo Janssen (ivo@ivo.nu)
Date: Sat Jul 22 2000 - 06:38:50 EST


Hi all,

We run a linux-fileserver that exports homedirectories to users.
The server runs 2.2.12 with usermode nfsd. "mount" says:
  tyrfing:/users/card/00 on /users/card/00 type nfs \
  (rw,bg,soft,rsize=8192,wsize=8192,addr=xx.xx.xx.xx)
Some of the clients are ultrasparcs running Solaris 2.5

The performance between linux and solaris is extremely bad, as has
been pointed out on this list a couple of times in the past.
The last time that I could find a thread on linux<->solaris
interoperatebility is from June 99. I hope that in one year
knowledge on the subject might have improved.

My questions:
1) How can this be? Is it the Solaris NFS implementation or the Linux
   NFS implementation?
2) Are there things to tweak on the solaris side?
3) Will another setup on the linux side improve things?
   Things that come to mind are:
   - tweak blocksize params
   - run 2.2 with kernelmode nfsd
   - run 2.3 or 2.4 (either userspace or kernelmode)

Hints very much appreciated!
(as is a CC to my emailaddress)

Regards

Ivo Janssen

--
+---------------------------------------------------------------------
| IVO JANSSEN - ivo at ivo.nu - http://ivo.nu/
| Delft University of Technology - the Netherlands
| finger ivo at server.ricardis.tudelft.nl for PGP and more info
| Part of the world's largest computer: http://www.distributed.net/

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Sun Jul 23 2000 - 21:00:18 EST