Since I'm running Weblogic RMI over HTTP, I fail to see
how an additional protocol layer over HTTP can achieve
faster speeds no matter it's efficiency.
As for using RMI directly (without HTTP), javasoft's
implementation does not scale well for large numbers of
concurrent clients, and neither javasoft or weblogic's
implementation will travel through a firewall.
(though weblogic's implementation is much more firewall
friendly since it uses a single port for it's traffic)
-David
-----Original Message-----
From: Cedric Beust [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, May 09, 2000 3:37 PM
To: [EMAIL PROTECTED]
Subject: RE: RMI/HTTP?
> From: A mailing list for Enterprise JavaBeans development
> This really doesn't make sense to me since using RMI
> (even an optimized version) should generate some
> additional overhead to the communication vs a raw
> URLConnection.
Why would that be ? RMI (well, ours anyway) pools connections and has an
optimized serialization mechanism. In constrast, HTTP will have to encode
the
MIME content, open and close the connection, etc...
I would bet that even Javasoft's RMI might actually be faster than HTTP
(going
out on a limb here). Did you try that as well ?
--
Cedric
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff EJB-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
