[OpenAFS] Extremely poor write performance.

[Neulinger, Nathan]

I've built it, and got interesting numbers... How dependent on other
parts of arla is that afsfsperf program - it might be worth seeing if it
can be broken out, as you have to go and also get kth-krb to get
afsfsperf built.

Alternatively, might be worth providing a set of afsfsperf binaries on
the openafs or grand.central.org sites. Then it becomes much more
reasonable to tell someone to "go run afsfsperf".

-- Nathan

------------------------------------------------------------
Nathan Neulinger                       EMail:  nneul@umr.edu
University of Missouri - Rolla         Phone: (573) 341-4841
Computing Services                       Fax: (573) 341-4216


> -----Original Message-----
> From: Jimmy Engelbrecht [mailto:jimmy@e.kth.se]=20
> Sent: Thursday, January 16, 2003 2:57 PM
> To: Robin Yamaguchi
> Cc: openafs-info@openafs.org
> Subject: Re: [OpenAFS] Extremely poor write performance.
>=20
>=20
> Robin Yamaguchi <rhy@physics.ucsb.edu> writes:
>=20
> > Hello All:
> >=20
> > I started working with openafs about a month ago, and i'm=20
> having problems=20
> > with write performance.  Please excuse any omitted=20
> information that is=20
> > important - i'm a newbie.  =3D)
> >=20
> > Here are some of the results from running bonnie++ (HD/filesystem=20
> > benchmarker) on an afs directory:
> >=20
> >      ------Sequential Output------ --Sequential Input- --Random-
> >      -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
> > Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
> > 496M  2351  24  2516   9   772   2  2793  25  3480   2  43.6   0
> >      ------Sequential Create------ --------Random Create--------
> >      -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
> > files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP /sec %CP
> >   16   159  26  2013  82   198  18   162  28  1013  70   215  15
>=20
> I dont think bonnie is a relevant benchmark for network filesystem.
>=20
> Have you tried something like:
>=20
> 1) Extract a tarball
> 2) compile a program
>=20
> Or maybe you should try the andrew benchmark, which is more=20
> relvant but
> very old.
>=20
> However if your are intressted in afs-performace-numbers you=20
> can compare
> with mine, statistics are done with afsfsparf from the=20
> arla-package. This
> is relevant when you test fileserver-performance, it does not
> performancetests client-performace.
>=20
http://www.e.kth.se/~jimmy/afsfsperf/afsfsperf.html

/Jimmy
_______________________________________________
OpenAFS-info mailing list
OpenAFS-info@openafs.org
https://lists.openafs.org/mailman/listinfo/openafs-info