BalancerMember returns 500 internal server error before failover

[Wayne W <waynemailinglists@xxxxxxxxx>]

Hi,

I'm seeing that with Apache 2.4.12 on Solaris when setup with mod_proxy to 2 tomcat instances when we bring down 1 node for maintenance the first client to try and access that node gets a Internal Server Error. Any requests after that (and for other sessions) are balanced correctly onto the other node.

Is this normal?

Is there a way for it to just balance over to the other node without returning the error?

My conf:

<VirtualHost *:80>
 ServerName somedomain.com
 ProxyPass /share http://127.0.0.1:8181/share

 ProxyPass / balancer://cluster/ stickysession=JSESSIONID nofailover=Off

 # define the balancer, with http and/or ajp connections
 <Proxy balancer://cluster/>
      Order allow,deny
      Allow from all
      BalancerMember ajp://127.0.0.1:8010 route=g1
      BalancerMember ajp://127.0.0.1:8011 route=g2
 </Proxy>
</VirtualHost>

I see in the apache logs the following:

Wed Apr 20 11:04:11.739700 2016] [proxy_ajp:error] [pid 58313] [client 10.12.32.251:56869] AH00992: ajp_read_header: ajp_ilink_receive failed, referer: https://example.com

[Wed Apr 20 11:04:11.739713 2016] [proxy_ajp:error] [pid 58313] (120006)APR does not understand this error code: [client 10.12.32.251:56869] AH00878: read response failed from 127.0.0.1:8010 (127.0.0.1), referer: http://example.com

[