Note: Handle big data that your business uses with dedicated hosting from NetHosting. We’re so confident that you’ll be satisfied, we offer a 30-day money back guarantee. Sign up today to get 24/7 support and a 100% uptime guarantee.
Added to the JBoss family of products is Data Grid 6 which will maybe make Red Hat a major player in the data center game.
Red Hat’s JBoss product line got an addition last month: Data Grid 6. Data Grid 6 is all about storage management within the middleware platform JBoss. The goal of the product is to give customers with e-commerce and financial transaction needs the ability to operate quickly and in an easy-to-scale environment.
Red Hat is quick to specify that this product is not like Hadoop or other big-data solutions. Data Grid is an in-memory and key-value store which is optimized to work with operations that systems like Hadoop can’t (e.g. those found in e-commerce and financial trading systems). Single transactions require a lot of reading and writing, which batch processing systems (e.g. Hadoop or other relational databases) can’t keep up with, and can’t scale as needed to make huge systems (like those at banks or trading houses or huge merchants like Amazon).
Computer scientist Eric Brewer came up with the Principles of Distributed Computing to address issues surrounding speed and scalability. The CAP Theorem of Brewer’s says that it’s not possible for a distributed computer system to provide consistency, availability, and partition tolerance simultaneously. The system can provide two of the three, but not all three. So, systems still face an uphill battle to try and provide at least two of these important factors.
Since Hadoop doesn’t fit the bill, Amazon uses non-relational Dynamo database to create an eventually consistent approach to their databases. There is a class of non-relational databases that is known as distributed key-value store (DKVS) and that is the class that Dynamo fits in. NoSQL is comprised of five classes, one of which is DKVS. Each of the five classes has a different architecture and approach to data management. There is just a key-value store database (KVS) that isn’t distributed across servers but instead is stored on disk or in RAM.
DKVS databases are also known as eventually consistent key-value store databases. They are designed to handle data that is spread over multiple servers. The databases accomplish this by using distributed hash tables for the key-value stores. Since they’re distributed, databases can use peer-to-peer connections between servers without a master control. The majority of the database options in this particular class are Dynamo-based implementations of Dynamo, like Project Voldemort, Dynomite, and KAI databases.
And that is the market and situation the JBoss Data Grid has entered, as of last month. The Data Grid product is based on the open source Infinispan platform, and the product is an addition to the JBoss Enterprise Application platform. The JBoss Enterprise Application platform is Red Hat’s bread-and-butter suite of middleware services. Infinispan was introduced in 2009 and was originally meant to be a data grid storage system optimized for the cloud. It was based loosely on JBoss Cache and it took the clustered-caching libraries of JBoss Cache to package together in a fully-equipped platform for the cloud.
Note: Get the ultimate flexibility and scalabilty in the cloud. With Cloud Virtual Hosting from NetHosting, you can experience the benefits of dedicated hosting at the price of shared hosting. Sign up today for unparalleled support and an unbeatable price.
The founder and project leader of JBoss Cache and Infinispan Data Grid is Manik Surtani. Now, three years later, his ultimate vision is getting its time to shine. The Data Grid software is accessible through a number of other applications like Representational State Transfer (REST), the memcache API, or the HotRod API. HotRod is another Infinispan spin-off that allows for the elasticity of Data Grid (i.e. the ability to scale in either direction as needed).
Manik Surtani explained that HotRod’s function allows two-way communication between clients and servers so when the shape of the data grid changes in the cloud, clients will still know where they fit within the dynamic data grid. That becomes important for efficiency with automated cloud systems.
If you’d like to read more about databases, check out the sneak peak we published about MySQL 5.6.