mod_cluster shows "All workers are in error state" when using manual ProxyPasses to /

Issue

• We have manually defined a / ProxyPass with mod_cluster:

  ProxyPass / balancer://mycluster/ stickysession=JSESSIONID|jsessionid nofailover=On
  

• We see the following occasionally logged in our error logs:

  [debug] mod_proxy_cluster.c(2193): proxy: byrequests balancer FAILED
  [error] proxy: CLUSTER: (balancer://mycluster). All workers are in error state
  

  But there seems to be no impact to our application requests.