home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ogicse!hp-cv!hp-pcd!news1.boi.hp.com!cupnews0.cup.hp.com!tardy.cup.hp.com!raj
- From: raj@cup.hp.com (Rick Jones)
- Newsgroups: comp.protocols.nfs
- Subject: Re: NFS I/O Ops/seconds
- Message-ID: <BrwKwn.4nD@cup.hp.com>
- Date: 24 Jul 92 17:07:35 GMT
- Article-I.D.: cup.BrwKwn.4nD
- References: <nkkvj9s@twilight.wpd.sgi.com>
- Distribution: usa
- Organization: The Place Where Packets Go To Die
- Lines: 58
- X-Newsreader: Tin 1.1 PL4
-
- coolidge@speaker.wpd.sgi.com (Don Coolidge) writes:
- : Seems like some summary is in order.
- :
- : 2) Sun (among others) has from time to time reported very
- : high IOPS numbers by changing the mix to have even fewer reads
- : and writes. This is not dishonest...it just means that the
- : reporting requirements for the test should make some mention
- : of the mix.
- :
-
- My understanding is that the number of reads and writes remained the same,
- just that the sizes are changed. This is particularly true if you try to
- make comparisons between nhfsstone 1.22 (which I beleive was the last of
- the nhfsstones that was supposed to be distributed) and nhfstone 2.0.X.
- (which really wasn't supposed to be distributed, but does get quoted by some
- folks...)
-
- : 4) LADDIS reccommends adding another Ethernet for every 200 IOPS.
- :
- : 5) However, that's not really necessary...our own tests have shown
- : that nhfsstones are at least as client-bound as they are server-bound
- : and media-bound. We've seen well over 400 IOPS over a single Ethernet
- : using nhfsstones with the default mix and with UDP checksumming
- : enabled, between a Crimson server and a single Crimson client. We
- : also see well over 500 IOPS with the same setup, but with two
- : Crimson clients on the single Ethernet. Multiple Ethernets
- : give an additional boost. So does use of FDDI instead of Ethernet
- : (a big boost).
- :
-
- 400 and 500 IOPS on a single Ethernet? Neat. I suppose that was nhfsstone
- 2.0.X, or were there modifications to the benchmark (read/write) sizes?
- Or were all the volumes mounted "private"?
-
- : 6) To be able to make reasonable NFS performance comparisons, I
- : think that at least the following need to be specified in the
- : performance report (there may well be more; this is just off the
- : top of my head):
- :
- :...
- : LADDIS is supposed to allow us to compare apples and apples. But it
- : seems to have gotten a tad too political to do so, as many of the
- : items I mentioned above were suggested and rejected as part of
- : the output report. Too bad.
- :
-
- Really? Which suggestsions were rejected? The list you gave sounded like
- the reporting requirements for LADDIS. Admittedly, I didn't make it to the
- last SPEC meeting, but all the previous meetings at which LADDIS was discussed,
- seemed to follow that thrust for reporting. Then again, having things happen
- at SPEC meetings is generally helped by attending. Some vendors don't get
- to the meetings. Too bad. Certainly happened the last time I didn't make it
- to a meeting.
-
-
- rick jones
- raj@cup.hp.com
- HP SPEC Rep for LADDIS
-