| Hello, I’m using Apache as a Load Balancer and have a question regarding the Balancer configuration on version 2.2.24 (Mac OS 10.8.5). This is my scenario: When some of the members in the group are down, sometimes the LB does not take the request and the client gets a timeout; if I remove those members from the config and restart Apache, all the requests are served (my configuration for the LB is below). Is my configuration missing something? maybe a retry or other timeout param? [=== START CONFIGURATION ===] <IfModule mod_proxy.c> ProxyPreserveHost On ProxyRequests Off ProxyReceiveBufferSize 2048 ProxyTimeout 300 Header add Set-Cookie "JSESSIONID=.%{BALANCER_WORKER_ROUTE}e; path=/" env=BALANCER_ROUTE_CHANGED ProxyPass /balancer-manager ! ProxyPass / balancer://cluster-ca/ ProxyPassReverse / balancer://cluster-ca/ <Proxy balancer://cluster-ca stickysession=JSESSIONID|jsessionid timeout=300 nofailover=Off> BalancerMember ajp://10.163.47.13:8009 loadfactor=1 route=ACIVRCAPRN01 ttl=60 BalancerMember ajp://10.163.47.14:8009 loadfactor=1 route=ACIVRCAPRN02 ttl=60 BalancerMember ajp://10.163.47.15:8009 loadfactor=1 route=ACIVRCAPRN03 ttl=60 BalancerMember ajp://10.163.47.16:8009 loadfactor=1 route=ACIVRCAPRN04 ttl=60 BalancerMember ajp://10.163.47.17:8009 loadfactor=1 route=ACIVRCAPRN05 ttl=60 BalancerMember ajp://10.163.47.18:8009 loadfactor=1 route=ACIVRCAPRN06 ttl=60 ProxySet lbmethod=byrequests </Proxy> <Location /balancer-manager> SetHandler balancer-manager Order Deny,Allow Allow from all AuthType Basic AuthName "Cluster manager " AuthUserFile /etc/apache2/.httpasswd Require valid-user </Location> </IfModule> [=== END CONFIGURATION ===] I appreciate any comment you can share with me. Regards, Israel Timoteo |
|