Chamara & Ishanni,
I suspect that this might be due to enabling of clustering, specifically,
state replication, amongst the AS nodes. This is because, once we restart
the AS instances, the throughput returns to normal. There is no need to
restart the LB, so the LB is behaving well. So, to test this hypothesis we
should run the following tests;

1. Setup an AS cluster, and send the requests from Java bench directly to
these nodes
2. Setup an AD cluster fronted with WSO2 LB (dynamic load balancing), but
disable state replication in the AS cluster
3. Setup AS without clustering, and front it with the WSO2 LB (static
endpoints)
4. Setup AS without clustering, and front it with an LB such as
mod_proxy_balancer

Run the same load tests with the same concurrency level for the above 4
setups, and compare the results.

-- 
*Afkham Azeez*
Director of Architecture; WSO2, Inc.; http://wso2.com
Member; Apache Software Foundation; http://www.apache.org/
* <http://www.apache.org/>**
email: **[email protected]* <[email protected]>* cell: +94 77 3320919
blog: **http://blog.afkham.org* <http://blog.afkham.org>*
twitter: **http://twitter.com/afkham_azeez*<http://twitter.com/afkham_azeez>
*
linked-in: **http://lk.linkedin.com/in/afkhamazeez*
*
*
*Lean . Enterprise . Middleware*
_______________________________________________
Carbon-dev mailing list
[email protected]
http://mail.wso2.org/cgi-bin/mailman/listinfo/carbon-dev

Reply via email to