I have completed some stresstests (with up to 500 concurrent users) on Apache
2.0.59, mod_jk 2.1.20, Tomcat 5.5.17, using the following
workers.properties:

<<<
# Define 1 real worker using ajp13
worker.list=ajp13

# Set properties for worker1 (ajp13)
worker.ajp13.type=ajp13
worker.ajp13.host=localhost
worker.ajp13.port=8009
worker.ajp13.connection_pool_timeout=600
worker.ajp13.connection_pool_minsize=10
worker.ajp13.connect_timeout=15000
worker.ajp13.prepost_timeout=10000
>>>

and this in the httpd.conf:

<<<
....
Timeout 300
KeepAlive On
MaxKeepAliveRequests 100
KeepAliveTimeout 15
# WinNT MPM
# ThreadsPerChild: constant number of worker threads in the server process
# MaxRequestsPerChild: maximum number of requests a server process serves
<IfModule mpm_winnt.c>
ThreadsPerChild 250
MaxRequestsPerChild 0
</IfModule>
....
>>>

and it seems to be coping just fine now under the stressload. I think I'll
use this setup on our live server now. Thanks a lot for your help.

Just one final question:

Despite my connection_pool_timeout=600 and connection_pool_minsize=10, one
of the Apache2.exe processes remains at 70MB memory usage, even after the
connection pool has gone down to 10 connections (verified by a netstat
call). The good news is, the memory doesn't seem grow any further though.
Still, keeping the 70MB for just the minimum 10 connections is a lot of
cache memory. Maybe the connection pool releases timed out connections as
expected, but doesn't flush them?


BTW.: The mod_jk 1.1.20, downloaded from
http://mirrors.dedipower.com/ftp.apache.org/tomcat/tomcat-connectors/jk/binaries/win32/jk-1.2.20/mod_jk-apache-2.0.58.so
identifies itself as mod_jk/1.2.19, not as mod_jk/1.2.20 !

Regards

J.Neuhoff


--
Sent from the Tomcat - User mailing list archive at Nabble.com.


---------------------------------------------------------------------
To start a new topic, e-mail: users@(protected)
To unsubscribe, e-mail: users-unsubscribe@(protected)
For additional commands, e-mail: users-help@(protected)