mod_cluster shows "All workers are in error state"

Solution Verified - Updated -

Issue

  • We can't reach JBoss through our mod_cluster load balancer.   mod_cluster is throwing errors saying "All workers are in error state"
  • Our proxied requests through mod_cluster are all yielding 503 responses from httpd.
  • The end user getting HTTP 503 errors and Apache httpd error_log shows below error.
proxy: CLUSTER: (balancer://mycluster). All workers are in error state
  • We have seen error All workers are in error state in the Apache httpd log during load test. The request finally failed with http 503 error. The Apache httpd error log show below error:
[debug] mod_proxy_cluster.c(2231): proxy: byrequests balancer FAILED
[error] proxy: CLUSTER: (balancer://mycluster). All workers are in error state
  • Modcluster - All workers are in error state.We often get this kind of errors in the logs file:
[error] proxy: CLUSTER: (balancer://mycluster). All workers are in error state

Environment

  • JBoss Enterprise Application Platform (EAP)
    • 5.x
    • 6.x
    • 7.x
  • JBoss Enterprise Web Server (EWS)
  • JBoss Core Services (JBCS)
    • Apache httpd
    • mod_cluster
      • 1.0.10.GA_CP0x
      • 1.2.x

Subscriber exclusive content

A Red Hat subscription provides unlimited access to our knowledgebase of over 48,000 articles and solutions.

Current Customers and Partners

Log in for full access

Log In