logrotate causes requests through mod_cluster to temporarily fail
Environment
- JBoss Enterprise Application Platform (EAP) 6.x
- mod_cluster
- Apache httpd
Issue
- We use logrotate, and when it rotates logs and restarts httpd processes, requests through mod_cluster temporarily fail
Resolution
- Use rotatelogs instead of logrotate to avoid the logrotate induced restarts
- If using logrotate, set
PersistSlots Onin your httpd mod_cluster config so that mod_cluster balancer states are persisted across restarts.- When using
PersistSlots, MODCLUSTER-427 can still cause short term request failures in newly initialized processes as the balancers aren't initialized in the new processes in time. Upgrade to mod_cluster 1.2.11.Final to address this.
- When using
Root Cause
- logrotate restarts httpd, which wipes its balancers by default as described in JBoss EAP unavailable after Apache restarts when using mod_cluster
This solution is part of Red Hat’s fast-track publication program, providing a huge library of solutions that Red Hat engineers have created while supporting our customers. To give you the knowledge you need the instant it becomes available, these articles may be presented in a raw and unedited form.
Comments