Hi,
I noticed start/stop cluster failures in sqe core tests.
I narrowed down the problem as below.
glassfish-3.1-b22.zip
http://sqe-hudson.sfbay.sun.com:8080/hudson/view/Sherry_v31/job/sherry-clu/
#6, OK to stop-cluster after a long sleep.
ant setup-cluster-profile
sleep 60
asadmin stop-cluster sqe-cluster
#7 failed to stop cluster with a short sleep.
ant setup-cluster-profile
sleep 5
asadmin stop-cluster sqe-cluster
+ asadmin stop-cluster sqe-cluster
remote failure: clustered_instance_1: Timed out waiting for
clustered_instance_1 to stop. clustered_instance_2: Timed out waiting
for clustered_instance_2 to stop. The command stop-instance failed for:
clustered_instance_1 clustered_instance_2 Command stop-cluster failed.
The setup-cluster-profile creates a cluster, sqe-cluster, of 2
instances on 1 machine.
The cli.log in hudson job gives details.
http://sqe-hudson.sfbay.sun.com:8080/hudson/job/sherry-clu/6/artifact/cli.log
http://sqe-hudson.sfbay.sun.com:8080/hudson/job/sherry-clu/7/artifact/cli.log
The stop-cluster failure caused more start/stop failures later.
Is this a product issue or a test harness issue?
Thanks for your advice.
Sherry