Cluster service still relocates when non-critical (__independent_subtree="2") fails to start in RHEL 5 or 6
Issue
- The
__independent_subtree="2"
option does not seem to take effect when a resource fails while the service is starting - When a
script
resource is marked as non-critical (__independent_subtree="2"
), and during service start the non-critical resource fails, it forces the service to failover to another node. By our expectation and understanding of non-critical attribute. the resource should have failed and service would be marked as partial leaving the other resources online on that initial node.
Environment
- Red Hat Enterprise Linux (RHEL) 5 Update 6 or 6 Update 1 or later with the High Availability Add On
rgmanager
- One or more resources marked with
__independent_subtree="2"
in/etc/cluster/cluster.conf
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.