home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.protocols.nfs:1939 comp.sys.sun.hardware:3515
- Path: sparky!uunet!pipex!unipalm!uknet!root44!root.co.uk!numb
- From: numb@root.co.uk (Matthew Newman)
- Newsgroups: comp.protocols.nfs,comp.sys.sun.hardware
- Subject: Re: NFS I/O Ops/seconds
- Message-ID: <numb.711892398@root.co.uk>
- Date: 23 Jul 92 11:53:18 GMT
- References: <1992Jul22.061146.15641@u.washington.edu>
- Sender: news@root.co.uk (News System)
- Organization: UniSoft Ltd., London, England
- Lines: 19
- X-Address: Spa House, Chapel Pl, Rivington St, London EC2A 3DQ
- X-Organisation: UniSoft Ltd
- X-Quote: To imagine is everything. - Albert Einstein
- Nntp-Posting-Host: dylan
- X-Job: Systems Engineer
- X-Phone: +44 71 729 3773 x215, Fax: +44 71 729 3273
-
-
- Assuming a typical mix (does such a thing exist?) a friend of mine
- and I did some very comprehensive NFS tests on a wide range of Sun
- and non-Sun equipment, a small paper was submitted to the Nice EUUG
- a few years ago.
-
- From memory our mix was 10-15% writes lots of getattr and lots of reads.
-
- The Sparc Servers at the time (SS2 type machines) with fast SCSI
- disks gave a sustained 80/sec, various machines can in the 40 -
- 60/sec range, the EPOCH NFS server gave 70/sec and wait for it:-
-
- The cheapest machine we tested: Sun SLC 86-92/sec. We spent six
- months doing this, and at the end bought 6 SLC's and 7GB of disk,
- I belive that they now have something like 15 SLC servers and about
- 20-30 GB of disk. They worked very well.
- --
-
- Matt Newman
-