Hi guys,

how to implement the memory-to-memory session replication in the WAS 7.0
environment.

we re trying to implement the M2M replication but no luck


Any help would be greatly appreciated.



Regards,
Brahmaiah

Hi Brahmi,

publib.boulder.ibm.com/infocenter/ieduas...

wpcertification.blogspot.in/2009/09/to-c...

Hope this could help in the process for implementation.

cheers. 

Hi, kasim

Thank you for your quick response.

We have successfully configured M2M settings in our production environment as per the below link. However failover is not happening in our environment. Whenever we are in the middle of some operation in our application, and if particular node which is serving the particular request goes down, session continuation is not happening instead  application is kicked out to our login screen again.


Regards,
Brahmaiah

Are you serializing the session information?  If not, the session information will not be stored in Memory..

hi..
can u provide more details on how have u configured them and what type of application is  that?

when the jvm1 goes down, and failover happens.. does the cookie value change - i mean the interchange of cloneid /partition id's?

Hello Ansar,

We have M2M enabled with replication mode "Both Client and Server Mode" and replication domain enabled for the cluster members with entire domain.

problem is that there is no failover happening in our environment. 

Also JSession in the webbrowser doesn't gets chaged when the moving to another jvm but i could see some additional values getting appended along with JSession ID.

for eg:
one of the scenario this was my JSession ID

JSESSIONID=0001jjHtCtdqPCLrmNe_Y3RPP5H:-3003HJ  

now if jvm goes down, my JSession ID is

JSESSIONID=0002jjHtCtdqPCLrmNe_Y3RPP5H:-3003HJ:1K4DR26OH3 

Hi All,
Please would like to know whether there is a way using wsadmin (JACL/Jython) script to do the real time session, dyna cache replication in between two nodes set for high availability but not clustered.
The scenario is that the application is kept on two nodes to keep them high availability but their is no deployment manager to cluster these nodes and hence to share the code in between them. Is there any way using scripting or other open source code to sync the code, session/dynacache replication in between these nodes in real time.
Kindly revert for any query. Thanks.