mod_cluster fails sending requests to Tomcat when using a connector executor

Issue

• We're running JBoss EWS Tomcat fronted by Apache httpd via mod_cluster. We can see from the logging and mod_cluster-manager page that httpd and tomcat are successfully connecting:

  Auto Refresh show DUMP output show INFO output
  Node worker1 (http://127.0.0.1:8371):
  Enable Contexts Disable Contexts
  Balancer: mycluster,Domain: ,Flushpackets: Off,Flushwait: 10000,Ping: 10000000,Smax: 1,Ttl: 60000000,Elected: 1,Read: 21645,Transferred: 451,Connected: 0,Load: -1
  Virtual Host 1:
  Contexts:
  
  /, Status: ENABLED Disable
  /manager, Status: ENABLED Disable
  /docs, Status: ENABLED Disable
  /examples, Status: ENABLED Disable
  /host-manager, Status: ENABLED Disable
  
  Aliases:
  
  localhost
  

• Requests through Apache/mod_cluster to Tomcat immediately fail with 503s, however, and the logging says all workers are in error state:

  [debug] mod_proxy_cluster.c(1253): proxy: byrequests balancer FAILED
  [error] proxy: CLUSTER: (balancer://mycluster). All workers are in error state