Complete stab in the dark here ... Tomcat serves a maximum number of threads at
a time (request processing threads). It may be that Tomcat's timings are
accurate for the threads *when* it executes them, but your HTTP Client will be
seeing a queue when Tomcat has no spare request thread
In my testbox, both of the following processes are running simultaniously:
java testprocess 1: 1000 threads (avg time to execute the
executeMethod:1723.61 ms)
java testprocess 2: 10 threads (avg resp time to execute the
executeMethod: 32 ms)
If it is to do with max processor attribute of
Hi,
Tomcat's max threads by default is usually around 75. I you fire 10 at it then
they all get serviced immediately.
If you send 1000, all 75 get taken up, meaning a queue forms for the 1000-75
remaining threads until Tomcat has spare threads for them.
Allistair
-Original
Hi,
I would think that is because Tomcat's max threads by default is usually around
75. I you fire 10 at it then they all get serviced immediately.
If you send 1000, all 75 get taken up, meaning a queue forms for the 1000-75
remaining threads until Tomcat has spare threads for them.
Oh I forgot to mention. Currently, the max thread size is 3000.
Note that both the client processes will run simultaniously. One with
more threads runs for long and one with less number of threads stops
quickly.
Moreover if tomcat runs out of threads then it will clearly put the info
in the