Each client directly connected to the server. Putting in an olbd seemed
somewhat useless since the redirection would happen only once and then the
olbd would be completely out of the picture. Also, we essentially know what
overhead (essentially a 1-time cost) the olbd adds to the flow so that can
As for the write benchmark, I agree. But, that is more difficult to devise.
Different files (more realistic but requires a lot more disk)? Write to
/dev/null? Things like that. I'll see what Bill Weels can come up with.
----- Original Message -----
From: "Fons Rademakers" <[log in to unmask]>
To: "Andrew Hanushevsky" <[log in to unmask]>
Cc: <[log in to unmask]>; <[log in to unmask]>;
<[log in to unmask]>; "Gerardo Ganis" <[log in to unmask]>;
"Fabrizio Furano" <[log in to unmask]>; <[log in to unmask]>;
"Alvise Dorigo" <[log in to unmask]>
Sent: Wednesday, February 02, 2005 11:01 PM
Subject: Re: Finally, the xrootd performance benchmark is here
> Hi Andy,
> very nice results indeed. Did each client go via the olbd or did they
> contact directly the xrootd server?
> Also, although xrootd is primarily a data server, it would be nice to see
> write benchmark, like a few hundred clients trying to open write streams
> the same time to a bunch of xrootd servers.
> Cheers, Fons.
> Andrew Hanushevsky wrote:
> > We finally got the xrootd performance benchmark together and put it on
> > web:
> > http://xrootd.slac.stanford.edu/presentations/Xrootd_Performance.htm
> > Frankly, we were suprised (pleasantly so, I add). The results came out
> > better that we expected.
> > Andy
> > -----------------------------------
> > "Algorithms from the NSA are considered a sort of alien technology: They
> > come from a superior race with no explanations."
> > --Bruce Schneier
> Org: CERN, European Laboratory for Particle Physics.
> Mail: 1211 Geneve 23, Switzerland
> E-Mail: [log in to unmask] Phone: +41 22 7679248
> WWW: http://www.rademakers.org/fons/ Fax: +41 22 7679480