home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.protocols.nfs
- Path: sparky!uunet!gatech!ukma!nntp.msstate.edu!news
- From: fwp@CC.MsState.Edu (Frank Peters)
- Subject: Re: NFS I/O Ops/seconds
- Message-ID: <1992Jul22.184738.16812@ra.msstate.edu>
- Sender: news@ra.msstate.edu
- Nntp-Posting-Host: jester.cc.msstate.edu
- Organization: Computing Center, Mississippi State University
- References: <1992Jul22.061146.15641@u.washington.edu> <64081@hydra.gatech.EDU>
- Distribution: usa
- Date: Wed, 22 Jul 1992 18:47:38 GMT
- Lines: 27
-
- In article <64081@hydra.gatech.EDU> gt1111a@prism.gatech.EDU (Vincent Fox) says:
- : In <1992Jul22.061146.15641@u.washington.edu> kint@rio.engr.washington.edu (Rick Kint) writes:
- : > We're buying a Sun to serve as an NFS server in one of our departments.
- : >The Sun rep commented that a typical Ethernet allows 300 NFS ops per second,
- : >so if you're on a single wire any server bandwidth beyond that is wasted.
- :
- : Hmmm. I'm a bit skeptical. All NFS operations are not created equal.
- : A write obviously takes a lot longer to complete than a getattr. I've seen
- : claims up to 1,000 or better. Auspex claims this I feel certain.
-
- Keep in mind that the 300 NFS ops per second refers to the number of
- operations for a single ethernet wire. Given a fairly usual mix of
- reads, writes, getattrs and lookups this is indeed a reasonable rough
- estimate of a maximum.
-
- If you run on a network with a higher bandwidth than ethernet then that
- maximum will be higher.
-
- If you have a server that will do 1,000 NFS ops/second and it is
- driving machines on four different ethernets then it can do more than
- 300 ops...but not on any one segment.
-
- So the significance of a such NFS limitations depends very much upon
- the nature and topology of your network.
- --
- Frank Peters - UNIX Systems Programmer - Mississippi State University
- Internet: fwp@CC.MsState.Edu - Phone: (601)325-7030 - FAX: (601)325-8921
-