tomcat-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Eric J. Pinnell" <>
Subject Re: TCP/IP issues with tomcat 4.1.24
Date Sat, 26 Jul 2003 00:00:51 GMT

That's interesting.  I haven't seen it but it's interesting.

Just curious, when the systems is exhibiting this behavior and you try to
bring up a browser and re-inact a testing scenerio what happens?  Is the
system still responsive?  Can you duplicate the error by hand?  When you
get that error what is the user experience?

I spend about as much time debugging testing tools as I do application
systems.  So I'm wondering what could be so different in 4.1.24 that would
cause that.  I think it's odd that it would goof up on the TCP because
that's handled at a much lower level.  It's also odd that the clients exit
out instead of chalking it up as an error and restarting themselves.  That
just might be a limitation of your testing tool though.

Check on the health of your testing tool (RAM, CPU) when it cracks up.

I don't know what kind of system you are using but make sure to check the
basics.  Make sure you have forced the systems (and the testing tools and
any switches in between) interfaces at 100 FDX (no auto-negogiate) and
tune the TCP settings.  That makes a world of difference.


On Fri, 25 Jul 2003, McClure, Timothy J(IndSys, GE Interlogix) wrote:

> We have a testsuite that sends 3 messages per second from 100 clients.
> With Tomcat version 4.1.18 this test runs for 24 hours.  With tomcat
> 4.1.24 the clients all eventually exit with a socket connect error after
> 2 hours.  Does anyone know why this might be happening?  Any help would
> be greatly appreciated
> Tim McClure
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message