[JBoss Data Grid] When Database is very expensive.
April 21, 2012
Imagine you have developed a successful web application built on traditional architecture which is visited by a consequent amount of users.
However, you are victim of you success, and more and more users are connected to you site.
Now, you are confronted to the overload of your application and your users tend to be unhappy due to the bad response time of your application.
Picture 2 :
So, in order to fix this you decided to change your application a little bit and then add some cache to it.
During a certain period of time, that clever trick worked. But, you still had the same problem.
Picture 3 :
Well, you had then the idea to increase the number of Application Servers to resolve that problem.
Nonetheless, with that decision you had to complexity your application. Nevermind, you had to increase the performance otherwise you were going to lose your customers.
But the problem still remained and you are now convinced the bottleneck is related to you old heavy database.
So you have no other choice than buy another database licence and its annual maintenance fees.
The MAJOR problem here is that is very very expensive.
Picture 5 :
Anyway, you decide adding more databases is not your only option.
You finally decide to implement a data grid in your environment in order to fix that problem without exploding your allocated budget.
If you are interesting by the subject then I am going to do a presentation/introduction about “JBoss Data Grid” in Belgium in April 2012.
If your company or you are based in BeNeLux and are interested by this presentation, just let me know and I will try to arrange a meeting for you.
N.B.: JBoss Data Grid 6 beta provides:
- A schema-less key value store: Red Hat JBoss Data Grid 6 beta is a NoSQL data store that provides flexibility to store any type of data.
- Compatibility with many types of applications: JBoss Data Grid 6 beta works with applications written in Java, C#, Spring, and others; deployed in an application server or standalone.
- Reliable grid-based data storage: Red Hat JBoss Data Grid 6 beta is designed to easily distribute data across multiple nodes. Data can be replicated in memory, written to disk, or stored in a relational database for fault-tolerance.
- Elastic scaling: JBoss Data Grid 6 beta allows simple addition and removal of storage nodes without disruption.
- Multiple access protocols: Red Hat JBoss Data Grid 6 beta provides easy access to the data grid using REST, memcached, Hot Rod, or simple map-like API.