I took a look at the code. UGLY. All of the code is in the JSP
and there are no class variables. There are lots of page scope
beans. Page scope is not something that I am used to using. I
usually use request. Could this be a problem where the next
request is hitting the page before it has
They are being posted to the same JSP. There are several beans
that are page scope. Think that is it?
I have a newer part that was written by me that uses request and
session scope beans. I will have the QA person run her tests
against that part of the project. I am assuming that this will
be
: I took a look at the code. UGLY. All of the code is in the JSP
: and there are no class variables.
Well, if it's all in one JSP, at least you have only one file to check. ;)
Sorry, couldn't resist...
: There are lots of page scope
: beans. Page scope is not something that I am used to
: machine 1 will sometimes get the results from query 2.
Smells like an instance/global variable where there shouldn't be one.
Add that to a subtle race condition and you have the problem you've
outlined.
Are the queries being posted to the same servlet?
-QM
--
software --
This has the taste of a thread safety issue. Do you guys happen to run the query
against the same servlet? If yes, do you happen to use variables, defined in the
servlet class scope in order to extract objects from session?
BR
-Original Message-
From: Norris Shelton [mailto:[EMAIL
Howdy,
Perhaps a threading bug that causes a race condition to the session
information. It could be in tomcat or in your webapp. It'll be hard to
come with a test case that always demonstrates this issue, but if you
can, please contribute the tiny webapp that illustrates the bug to
bugzilla.
