Satyajit, et al:
Just closing the loop in case others run into this same issue....
I have installed and configured the latest Sun Glassfish Enterprise
Server (sges_ee-2_1-p01-linux.bin) and my problem was resolved.
pgs
Satyajit Tripathi wrote:
> Hello Paul,
>
> As I have come across, there are few known cluster bugs (with 2 systems
> setup) identified with the open source version of GlassFish V2.1 on
> Linux. Those have been fixed in GlassFish Enterprise Edition that comes
> with the GUI Installer, and is not JAR based. Please see the Document-id
> 128642-09 in public domain
> http://sunsolve.sun.com/search/document.do?assetkey=1-21-128642-09 for
> more details.
>
> You may need to download and install the GlassFish Enterprise Server
> V2.1 Patch01 (i.e. SJS Application Server 9.1_02 Patch07) to test.
>
> Please let me know if the information was useful to you ?
>
> Thanks & regards
> --Satya
>
>
> Paul G. Szabady wrote:
>> Greetings,
>>
>> I'm a bit new to glassfish and have had some luck with setting up a
>> single server cluster using one node agent and 4 instances. Taking
>> this a step further, I'm trying to double the set up with a second
>> server, 2 nodeagents, and 8 instances.
>>
>> Both RHEL 4.7 servers have glassfish v2.1 installed using the
>> setup-cluster.xml. Server1 is the DAS. Both server1 and server2 have
>> nodeagents running on them and can be seen through the admin gui:14848
>> on the DAS. Server1 has four instances configured and running.
>>
>> Here's the issue. When I create an instance on server2 it shows up as
>> a stopped instance (in both the gui and through asadmin), but it won't
>> start. I believe it won't start because the directory hierarchy for
>> the instance (ie: $installpath/nodeagents/node-agent-2/new-instance)
>> doesn't get created.
>>
>> Any suggestions?
>>
>> The commands I used to create/start the node agent and instance on the
>> second server
>> - for the node agent:
>> /usr/local/gf/bin/asadmin create-node-agent --user admin --host
>> 192.168.2.8 --port 14848 node-agent-2
>>
>> /usr/local/gf/bin/asadmin start-node-agent --user admin node-agent-2
>>
>> - for the instance:
>> /usr/local/gf/bin/asadmin create-instance --user admin --host
>> 192.168.2.8 --port 14848 --cluster cluster --nodeagent node-agent-2
>> --systemproperties
>> "JMX_SYSTEM_CONNECTOR_PORT=8692:IIOP_LISTENER_PORT=3332:IIOP_SSL_LISTENER_PORT=4442:IIOP_SSL_
>> MUTUALAUTH_PORT=5552:HTTP_LISTENER_PORT=1112:HTTP_SSL_LISTENER_PORT=2222"
>> instance2
>>
>> I've also tried to create the instance using the admin gui and
>> received the same results: instance gets created, but no supporting
>> directories so the instance will not start. The nodeagents on both
>> servers have been restarted using "--syncinstances=true" as well, but
>> to no avail.
>>
>> Odd numbered instances reside on server1, even on server2.
>>
>> /usr/local/gf/bin/asadmin list-instances --user admin --host
>> 192.168.2.8 --port 14848
>> instance1 running
>> instance3 running
>> instance5 running
>> instance7 running
>> instance2 not running
>> Command list-instances executed successfully.
>>
>> Final note: iptables are currently configured to allow all traffic
>> to/from server1/server2.
>>
>> Any suggestions or assistance would be appreciated!
>>
>> pgs
>>
>
>