It is possible to get the effect of clustering without involving the entire
container. This effectively involves making the application itself clustered
in the following way using queues.
The way to do this is as follows.
a> Create several copies of a flow with a JMS Consumer BC as the first
JBI Service/Endpoint and associate it with the same queue.
b> Make sure the individual containers on which the flow copies are
hosted on are SEDA
Note that this way of clustering flows is
a> at the application layer
b> the payload being moved on the queue are messages as opposed to
message exchanges.
c> there are no JMS queues that are associated with each unique JBI
Service Name/Endpoint Name.
In effect what you get is a kind of poor man clustering capability afforded
by the application rather than infrastructure layer.
The benefits are are also significant in that since the payload is a message
as opposed to Message Exchange, the MEP related interactions between JBI
Service and endpoints are eliminated leading to a better performance profile
while getting the benefit of clustering
Hope this helps.
- Ashwin
-----
---
Ashwin Karpe, Principal Consultant, PS - Opensource Center of Competence
Progress Software Corporation
14 Oak Park Drive
Bedford, MA 01730
---
+1-972-304-9084 (Office)
+1-972-971-1700 (Mobile)
----
Blog: http://opensourceknowledge.blogspot.com/
--
View this message in context:
http://www.nabble.com/clustering-without-containers--tp20619081p20646521.html
Sent from the ServiceMix - User mailing list archive at Nabble.com.
