Details
-
Bug
-
Status: Resolved
-
Critical
-
Resolution: Cannot Reproduce
-
cellar-2.2.4
-
None
Description
The scenario is the following:
- Start up an instance of Karaf
- Install cellar (using default hazelcast broadcast enabled config)
- Join a new group
- Quit the default group
- Verify that the default group has been left using cluster:group-list ( sometimes, your node will still be in default group even though the quit command gives a different feedback!!! ) - if this fails, quit the default group again, until your node really appears to be out of the default group
- Start a second instance of Karaf
- Install cellar (using default hazelcast broadcast enabled config)
- Verify the configuration of the second container. In my case I realized that there were configurations pushed from the first container to the second, e.g. in Pid: org.ops4j.pax.logging I found paths from the first container inside the configuration of the second container (log4j.appender.sift.appender.file, log4j.appender.out.file)