tomcat-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From bugzi...@apache.org
Subject DO NOT REPLY [Bug 33510] - recover_wait_time for workers in_error_state should be calculated using difftime()
Date Sat, 12 Feb 2005 00:23:53 GMT
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG·
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://issues.apache.org/bugzilla/show_bug.cgi?id=33510>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND·
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=33510





------- Additional Comments From rainer.jung@kippdata.de  2005-02-12 01:23 -------
Maybe this is not the right place to go deeper into it, tomcat-user is better.
As a final point inside bugzilla I would be interested, if you could confirm or
deny my explanations by checking the log file.

Before we switch over to tomcat-user:

1) Mladen Turk is active in including global worker status fr mod_jk even for
the multi process model by using shared memory. It might take some time to get
stable, but then all processes will have a global view about error status and
handled load for any worker.

2) To disable sending requests to a tomcat server, you could delete it from the
balancer list and gracefully restart apache (but sometimes processes might hang
if you do it under very high load).

3) To enable more graceful shutdown and if you are using sessions with
stickyness, you could give all workers the local flag and then before you are
going to maintain some tomcat you remove the local flag. Then it will only
handles requests for sessions that have been started on this instance and not
get any new sessions.

4) Even in case of network failures you might be able to get a much faster error
detection by using the CPING/CPONG feature and setting recovery_opts and
adequate timeouts. Study the new mod_jk documentation, there are various
timeouts. But be careful not to make the timeouts too short. Depending on the
response times of your application timeouts between 20 and 120 seconds might
adequate. You can also experiment with lowering the retry count.


-- 
Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

---------------------------------------------------------------------
To unsubscribe, e-mail: tomcat-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: tomcat-dev-help@jakarta.apache.org


Mime
View raw message