Hi,

In our environment we have 5 nodes each having 14 jvms in a single cluster.

WAS 6.1 on AIX.

we have minor updates on app.so after updating app jvms need to be restarted

inorder to take effect.

is there any method to restart each node one after another inorder to get High Availability??

Thanks
Ravi 

Hi Ravi,

  If you have a HTTP server with plugin in front of your cluster, you can do a RippleStart, the cluster members stops and starts in order and the plugin balance the request to other members.

 http://publib.boulder.ibm.com/infocenter/wasinfo/v6r1/topic/com.ibm.websphere.nd.doc/info/ae/ae/trun_wlm_cluster_start.html

  RippleStart combines stopping and starting operations. It first stops and then restarts each member of the cluster.


  Hope this helps

regards,

Hi Ravi,


  Looking other questions i have see that you ask something similar in

www.websphereusergroup.org/go/thread/vie...

  If you want control how and when stop the nodes you need to do and script for examples with semaphores testing tha availability of the node, sending the stop command to each server and when they start continue with the other members.

  take care about the wsadmin invocation method SOAP/RMI timeouts the files are located at:

  \profiles\\properties\

   Soap.client.props
com.ibm.SOAP.requestTimeout=180

  sas.client.props

  com.ibm.CORBA.requestTimeout=180 

hope this helps,

regards,

I think if you want HA for an applications, you need to use merge plugin concept, where you can have all the nodes with their respective jvms/apps. When you stop/start one node, plugin will route request to other node.