does anyone knows how to configure the mod proxy balancer method in order to balance on a secondary backend instance only in case of an overflow of simultaneous requests?
More specifically:
I have an Apache + mod proxy balancer which is in front of a primary Tomcat instance (with https connector) which is able to serve only N simultaneous requests, if it receive more requests it crashes or start responding slow.
I prepared another secondary remote tomcat instance which will be used only for the N+1 requests, if there are less than N simultaneous request it should not be used at all.
I would like to tell Apache to balance on the primary Tomcat instance to max N simultaneous requests. The N+1 request must be balanced on the secondary remote Tomcat instance.
I already tried to use:
- status=+H
on the secondary member and max=N
on the primary member on BalancerMembers settings
- maxThreads=N
and acceptCount=N
on the primary tomcat HTTPS connector
but it is not working as expected.
If i test a heavy load i can see Apache not forwarding to secondary the N+1 requests, but it is forwarding all of them on the primary.
Anyone may help me?
Thank you very much.