You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tomcat.apache.org by Bill Ataras <ba...@mobilemob.com> on 2000/01/23 01:06:12 UTC

Design issues: mod_jserv or mod_rewrite[P] or all of above?

I've been playing with latest tomcat (3.1m2). Looks like it reloads jsps
now. yay! And mod_jserv seems to work fine with virtual hosts. (using
apache 1.3.9).

I haven't been able to get a rewrite [PT] rule to shove the new URI back
through jserv to tomcat running on internal machine. Using rewrite [P]
to simply proxy the request to tomcat and bypass jserv all together
seems to work great.

So which would stress the system less?

1. URI -> appjservmount -> tomcat
2. URI -> rewrite -> PT -> appjservmount -> tomcat
3. URI -> rewrite -> tomcat:8080

I haven't been able to get #2 to work. 1 & 3 are fine. I would assume
that if you HAVE to use a rewrite, you'd want it to go through jserv
(#2)? Presumably jserv mux's requests over a single socket and will
eventually offer load balancing?

Re: Design issues: mod_jserv or mod_rewrite[P] or all of above?

Posted by Jean-Luc Rochat <jn...@cybercable.fr>.
Bill Ataras wrote:
> 
> I've been playing with latest tomcat (3.1m2). Looks like it reloads jsps
> now. yay! And mod_jserv seems to work fine with virtual hosts. (using
> apache 1.3.9).
> 
> I haven't been able to get a rewrite [PT] rule to shove the new URI back
> through jserv to tomcat running on internal machine. Using rewrite [P]
> to simply proxy the request to tomcat and bypass jserv all together
> seems to work great.
> 
> So which would stress the system less?
> 
> 1. URI -> appjservmount -> tomcat
> 2. URI -> rewrite -> PT -> appjservmount -> tomcat
> 3. URI -> rewrite -> tomcat:8080
> 
> I haven't been able to get #2 to work. 
Try again. I did it, and it worked. but I don't have the rewrite rules
here.

> 1 & 3 are fine. I would assume
> that if you HAVE to use a rewrite, you'd want it to go through jserv
> (#2)? 
see above.

> Presumably jserv mux's requests over a single socket 
no, there is one socket/httpd proc
> and will
> eventually offer load balancing?

yes (at least with JServ for the moment).

Jean-Luc