Setup: minas and apollo are large (8 GB RAM multicore) x80 machines
running OpenSolaris (build 134b I think).
This is with JDK 1.6.0_16.
I am attempting to creating a simple 3 instance cluster as follows:
asadmin start-domain
asadmin create-cluster c1
asadmin --user admin create-node-ssh --nodehost minas --installdir
/volumes/work/GlassFish/v3/glassfish3 minasNA
asadmin --user admin create-node-ssh --nodehost apollo --installdir
/volumes/work/GlassFish/v3/glassfish3 apolloNA
asadmin create-instance --node minasNA --systemproperties
instance_name=in0 --cluster c1 --portbase 9000 --checkports=true in0
asadmin create-instance --node apolloNA --systemproperties
instance_name=in1 --cluster c1 --portbase 10000 --checkports=true in1
asadmin create-instance --node minasNA --systemproperties
instance_name=in2 --cluster c1 --portbase 11000 --checkports=true in2
asadmin set configs.config.c1-config.java-config.debug-enabled=true
asadmin start-cluster c1
(note that minas and apollo can both ssh to each other without
passwords: all the public keys
are properly configured.)
start-cluster fails to create instance in1, and prints the following:
* in1: Could not start instance in1 on node apolloNA (apollo).
Command failed on node apolloNA (apollo): DAS host localhost does
not match DAS host used by node apolloNA.
Node apolloNA is configured to connect to DAS on minas:4848 with
secure = false.
Command start-local-instance failed.
To complete this operation run the following command locally on
host apollo from the GlassFish install location
/volumes/work/GlassFish/v3/glassfish3:
asadmin start-local-instance --node apolloNA --sync normal in1
The command start-instance executed successfully for: in2 in0
The command start-instance failed for: in1
Command start-cluster completed with warnings.*
I need to create the instance from minas, since that's where the test is
running. This is fairly important
to my debugging of a number of IIOP FOLB issues.
This has worked until recently, so what changed this time?
What other command(s) do I need? I tried
asadmin enable-secure-admin
(which I don't understand at all, even after reading the command help)
This does not fix this problem, and causes other things to fail (like
starting the other instances,
and weird problems with deployment).
If this is RTFM, please point me to the FM so I can RT.
Thanks,
Ken.