Jung , Dr. Christoph schrieb:
You should in no case for your own sake meet a thus serious design decision
without looking at the exact reasons.
E.g., there was someone mentioning that latencies in the area of a second
could be caused by the
.NET SOAP library trying to retrieve proxy settings with each new http
connection! Hence 90% of the call would be consumed without even touching a
part of SOAP.
CGJ
So, do you think in the end SOAP should be as fast as IIOP? Can any of the authorization settings be tuned in jboss-net.sar to enhance the throughput (no authorization required..)? I am not yet convinced that our experience of dramatically decreasing performance while increasing the number of concurrently accessing clients using soap clients versus RMI-IIOP clients is not caused by the design of the soap protocol. But to be sure that no other effects like proxy settings make things worse I have to examine all clients. Can�t do it at the moment unfortunatly...
Thanks, Best regards, Carsten
------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ JBoss-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/jboss-user
