While running certain scenarios during testing of the Carbon 4.2.0 packs, we saw certain performance issues when distributed Hazelcast HashMaps are used as the backing data structure. To overcome this, I have introduced the concept of an L1 local cache. It is a write through cache, where items stored in it will also get written to the distributed data structure. When there is a L1 cache miss, we will check in the L2 distributed cache to see whether the value is there.
I am seeing significant performance improvement in the tenant creation scenario after this improvement. -- *Afkham Azeez* Director of Architecture; WSO2, Inc.; http://wso2.com Member; Apache Software Foundation; http://www.apache.org/ * <http://www.apache.org/>** email: **az...@wso2.com* <az...@wso2.com>* 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*
_______________________________________________ Dev mailing list Dev@wso2.org http://wso2.org/cgi-bin/mailman/listinfo/dev
