> With Tomcat and servletexec, the 3 results of the 3 instances are
> differents, and it seems that the 3 instances or at least 2 are executed in
> parallel..
> With resin 2.10, the 3 instances are exactly executed in sequence so the 3
> results are good and equals..
>
> I change nothing between the tests except the servlet container ..
>
> bizarre ...

Not really. If you study that bit of the specification (src 2.2.1 in
the 2.3 spec) you'll note that pooling of STMd servlets is
optional. It seems that ServletExec has not got a pooling system.

ServletExec isn't the only container without a pooling system for STM
servlets, GNU-Paperclips doesn't have one either, largely because I
don't believe in STM.


Nic Ferrier

___________________________________________________________________________
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff SERVLET-INTEREST".

Archives: http://archives.java.sun.com/archives/servlet-interest.html
Resources: http://java.sun.com/products/servlet/external-resources.html
LISTSERV Help: http://www.lsoft.com/manuals/user/user.html

Reply via email to