Hi Ingo,
Ingo Düppe wrote:
Am I'm right, that with the following setting a reply_timeout is
detected when the tomcat doesn't reply within 10 seconds and it will be
set into error if more than 15 timeouts in a minutes are detected?
worker.plexus.reply_timeout=10000
Yes. The timeout gets reset every time mod_jk waits for the next
package. So it's not an overall response time timeout, but it limits the
longest pause allowed between reposnse packets. Usually it will fire,
between the point in time the request was forwarded to Tomcat and the
first request packet arrives. Most apps don't have huge delays after
that point.
See also
http://tomcat.apache.org/connectors-doc/generic_howto/timeouts.html
for a description of all possible timeouts.
worker.plexus.max_reply_timeouts=30
In case you assume a steady rate of long running requests, yes: 15 per
minute. In case there were no long running requests for a long time, our
intrnal counter will be zero and we will allo up to 30 long running
requests before triggering the error status.
The counter gets divided by 2 every minute, so that older events count
less the longer they are away. A worker that worked fast for a long time
can use all of the 30 events, if it already had 30 (or 29) timeouts in
the minute before, only 15 are free the next minute.
Regards,
Rainer
---------------------------------------------------------------------
To start a new topic, e-mail: users@tomcat.apache.org
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
