I'll have to look into whether there is a workaround for that
ThreadLocal bug, and figure out a way to test this.

On Jan 8, 2008 2:08 PM, jason lea <[EMAIL PROTECTED]> wrote:
> Hi,
>
> We get this error quite often with Java 1.5, we asked about it in the
> mailing list back in August.  We found the error doesn't occur in Java 1.6.
> We usually see a stack trace with a NPE when doing getSession() or
> getCookies().
>
> Here is what I wrote last time:
>
> We are getting an intermittent NullPointerException in Tapestry 5.0.5.  We
> are running on Java 1.5.
> It seems that the RequestGlobals has a null Request when the call to
> getCookies is made (or that RequestGlobals is null).  We cannot get the
> error to occur every time so testing is tricky, sometimes it occurs on the
> first request after the webapp starts with subsequent requests good for a
> few minutes before it occurs again.  Other times it starts fine and pressing
> F5 100 times might make it occur.
>
> Our thought at the moment is that a ThreadLocal bug (
> http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6550283 ) might be
> causing the RequestGlobals to initially be correctly initialised and stored
> in ThreadLocal but later on with more items being added this bug means
> further down the track, a new RequestGlobals is added with null values.
>
>
>
> --
> Jason Lea
>



-- 
Howard M. Lewis Ship

Creator Apache Tapestry and Apache HiveMind

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to