tomcat-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rainer Jung <rainer.j...@kippdata.de>
Subject Re: Mod_jk load balacing algorithm
Date Mon, 07 Aug 2006 23:29:05 GMT
Hi,

Mohan Wickramasinghe schrieb:
> Hello
> 
> We are using the "new" Busyness method with 1.2.18
> We have also using 'sticky sessions'.
> 
> Clarification on
> "> As soon as there will be real load, behaviour will differ."
> 
> We have 9 nodes on 3 identitical servers with identical settings, lb
> factor equal on all nodes, but we see one node which had got a large MAX
> value (due to a large Busy number at some point), and one node that is not
> getting hardly any requests.
> 
> Why such deviations please with "Busyness" ?
> Is it or could it be due to using 'sticky sessions' ?
> 

Hard to tell without any numbers. Stickyness is somewhat contrary to
load-balancing, but when the number of sessions grows, one would
statistically expect the load to get even.

You could increase JkLogLevel and add request logging to find out ore
about the balancing decisions.

Regards,

Rainer

---------------------------------------------------------------------
To start a new topic, e-mail: users@tomcat.apache.org
To unsubscribe, e-mail: users-unsubscribe@tomcat.apache.org
For additional commands, e-mail: users-help@tomcat.apache.org


Mime
View raw message