Trouble w/Session Management on 4.1.29

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Trouble w/Session Management on 4.1.29

Craig A. Vanderborgh
Hello:

We have and embedded version of Jetty 4.1.29 running on Windows CE.  I
know - running anything on Windows CE, much less a Java application
server, is an ill-advised thing to do.  Sadly that can't be helped.

Our version of Jetty is compiled to native code using our own GCJ
compiler and runtime.  It actually works pretty well.  But we do have a
problem that I need to ask about.  The problem is with HTTP session
management.  Occasionally, the jsessionid of a given client/server
interaction will suddenly CHANGE, even though the session timeout hasn't
been exceeded.  Naturally, this causes some problems.  Even worse, we've
encountered a problem where the session appears to timeout even though
there's no way that the configured session timeout value could have been
exceeded.  Naturally, the exact same webapp runs flawlessly on the
embedded Linux version of our gcj-jetty.

So my question is really just an archeological one - where in the Jetty
source hierarchy is the session timeout stuff dealt with, and might
someone have a few tips on how to go about investigating these HTTP
session anomalies?  I am sure that there's nothing wrong with Jetty, the
problem is almost certainly a CE/GCJ issue in our infrastructure in some
way.  I'm just looking for a few kind words on how best to proceed with
effectively debugging this HTTP session stuff in Jetty.

Thanks in advance,
craig vanderborgh
voxware incorporated


-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
jetty-discuss mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/jetty-discuss
Reply | Threaded
Open this post in threaded view
|

Re: Trouble w/Session Management on 4.1.29

Chris Haynes

org.mortbay.jetty.servlet.HashSessionManager

is a good place to start.

Is you problem anything to do with Long handling?  Is Long addition working OK -
when there is a 'carry' crossing the bit 31/32 boundary?

Another stone to overturn: Are the low-level hashcode()  methods working OK - on
Long and on String?

Chris Haynes


"Craig A. Vanderborgh" asked

> Hello:
>
> We have and embedded version of Jetty 4.1.29 running on Windows CE.  I know -
> running anything on Windows CE, much less a Java application server, is an
> ill-advised thing to do.  Sadly that can't be helped.
>
> Our version of Jetty is compiled to native code using our own GCJ compiler and
> runtime.  It actually works pretty well.  But we do have a problem that I need
> to ask about.  The problem is with HTTP session management.  Occasionally, the
> jsessionid of a given client/server interaction will suddenly CHANGE, even
> though the session timeout hasn't been exceeded.  Naturally, this causes some
> problems.  Even worse, we've encountered a problem where the session appears
> to timeout even though there's no way that the configured session timeout
> value could have been exceeded.  Naturally, the exact same webapp runs
> flawlessly on the embedded Linux version of our gcj-jetty.
>
> So my question is really just an archeological one - where in the Jetty source
> hierarchy is the session timeout stuff dealt with, and might someone have a
> few tips on how to go about investigating these HTTP session anomalies?  I am
> sure that there's nothing wrong with Jetty, the problem is almost certainly a
> CE/GCJ issue in our infrastructure in some way.  I'm just looking for a few
> kind words on how best to proceed with effectively debugging this HTTP session
> stuff in Jetty.
>
> Thanks in advance,
> craig vanderborgh
> voxware incorporated
>
>
> -------------------------------------------------------
> SF.Net email is Sponsored by the Better Software Conference & EXPO
> September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
> Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
> Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
> _______________________________________________
> jetty-discuss mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/jetty-discuss
>
>
>





-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
jetty-discuss mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/jetty-discuss
Reply | Threaded
Open this post in threaded view
|

Re: Trouble w/Session Management on 4.1.29

Greg Wilkins-5
In reply to this post by Craig A. Vanderborgh
Craig A. Vanderborgh wrote:
> Hello:
>
> We have and embedded version of Jetty 4.1.29 running on Windows CE.  I
> know - running anything on Windows CE, much less a Java application
> server, is an ill-advised thing to do.  Sadly that can't be helped.

Wow - I'm impressed.

But again, I think I'd recommend looking at Jetty 6 to future proof
yourself.  It is much simpler than even Jetty 3 and I'll be glad to
help make any changes to make GCJ work on it.

As for 4.1.29, Chris has pointed you in the right direction.

It would also be good if you could log the situation when a
new session is created.  Was an old session ID presented that
was not found (or found an invalid session).  Was no session ID at
all presented?

cheers


> Our version of Jetty is compiled to native code using our own GCJ
> compiler and runtime.  It actually works pretty well.  But we do have a
> problem that I need to ask about.  The problem is with HTTP session
> management.  Occasionally, the jsessionid of a given client/server
> interaction will suddenly CHANGE, even though the session timeout hasn't
> been exceeded.  Naturally, this causes some problems.  Even worse, we've
> encountered a problem where the session appears to timeout even though
> there's no way that the configured session timeout value could have been
> exceeded.  Naturally, the exact same webapp runs flawlessly on the
> embedded Linux version of our gcj-jetty.
>
> So my question is really just an archeological one - where in the Jetty
> source hierarchy is the session timeout stuff dealt with, and might
> someone have a few tips on how to go about investigating these HTTP
> session anomalies?  I am sure that there's nothing wrong with Jetty, the
> problem is almost certainly a CE/GCJ issue in our infrastructure in some
> way.  I'm just looking for a few kind words on how best to proceed with
> effectively debugging this HTTP session stuff in Jetty.
>
> Thanks in advance,
> craig vanderborgh
> voxware incorporated
>
>
> -------------------------------------------------------
> SF.Net email is Sponsored by the Better Software Conference & EXPO
> September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
> Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
> Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf



-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO
September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices
Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA
Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
_______________________________________________
jetty-discuss mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/jetty-discuss