W3C home > Mailing lists > Public > www-jigsaw@w3.org > September to October 1996

Caching - how?

From: Anselm Baird-Smith <abaird@w3.org>
Date: Thu, 19 Sep 1996 09:13:26 +0500
Message-Id: <9609191313.AA04422@www18.w3.org>
To: gs@christine.zfe.siemens.de (Gernot Schober)
Cc: www-jigsaw@w3.org

[Gernot, the mailing list, cc'ed in the reply is www-jigsaw@w3.org]

Gernot Schober writes:
 > Hi everybody!
 > 
 > I am quite a newbie to the java-language, but I am very interessted in how
 > the Jigsaw server can be *so* fast.

Yeah ! thanks, remind me to pay you a beer for that one :-)

 > I were told, that java is not the very fastest language at runtime, so I
 > wondered for long time, how it works, that the requests are so fast
 > responded.
 > 
 > The only (not very sattisfying) answer I found until today is, that the
 > Jigsaw does some kind of caching - both working in the proxie mode and
 > working as a real server. What i am now interessted in is, how is that
 > caching implemented (and where in the src can i find it)?
 > Up to now i found that caching is part of a (client-sided?) filter, but
 > I can not find that anywehre.

I don't know what version of Jigsaw you are using (probably
1.0alpha1), and against what you compared it. If you compare Jigsaw to
the CERN server, then no doubt it seems fast: the CERN server forks a
new process off on each request.

The only thing Jigsaw caches is the meta-information associated to the
objects (ie "resources") it serves. In 1.0alpha2, it does a little bit
more then just that, but it doesn't cache (yet) the content of the
resources in memory. In general, 1.0alpha2 is really faster then
1.0alpha1, in particular under heavy load.

Anselm.
Received on Thursday, 19 September 1996 09:13:47 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 9 April 2012 12:13:25 GMT