On 09.12.2009 12:18, Pid wrote:
On 09/12/2009 11:00, Alessandro Bahgat wrote:
Nishant,
that didn't quite work, actually.
After some struggle with our outsourcers, we added a new machine
running Tomcat 6.0.20 *without tcnative* (they misplaced the .so
files).
That eventually resulted in a worse behaviour: within a few minutes
from startup all the TP-Processor threads on that machine get stuck on
the java.net.SocketInputStream.socketRead0 method.
We are about to test the same server using
- Tomcat 6.0.20 + tcnative 1.1.18
- Tomcat 6.0.16 + tcnative 1.1.18
(Our original configuration runs on Tomcat 6.0.16 + tcnative 1.1.10)
In the meanwhile, we found out these lines in the mod_jk log:
[Thu Dec 03 16:58:52 2009][31539:42688] [info]
service::jk_lb_worker.c (873): unrecoverable error 200, request
failed. Client failed in the middle of request, we can't recover to
another instance.
Could that be related to a firewall not working properly between
Apache and Tomcat?
It could be, but while you're upgrading you might consider upgrading
HTTPD to the best available version too, 2.0.52 release date: 1 Oct
2004. (That's 35 internet years ago.)
... and mod_jk 1.2.19 dates back to September 2006, so according to your
math 21 internet years ago. There were so many bugs fixed since then,
that you'll hardly find anyone that really tries to help debugging those
old versions. At least use a recent mod_jk and look at
http://tomcat.apache.org/connectors-doc/generic_howto/timeouts.html
for some important hints about configuration.
Regards,
Rainer
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org
