New application of Clustrix Sierra in cloud era

With the development of cloud computing, more and more data development work has been developed on cloud platform.

Clustrix, a start-up Newsql database company, recently received a 16.5 million-dollar C-round financing that includes Sequoia Capital, U.S Venture, and ATA Ventures. After that, the innovation company, founded in 2005, has a total financing amount of $46.5 million trillion. Clustrix, headquartered in San Francisco, is a start-up project in 2006 funded by Y Combinator. Now the momentum of the rapid development is afraid that Y Combinator had not expected.

Clustrix Sierra in the industry is called the cloud computing era of MySQL. Two years ago, the company launched a highly scalable telescopic database solution Sierra, which provides functionality similar to the SQL database, while extending the data store.

"We've run our apps on more than one cloud platform in the world, but now our goal is to open this distributed system to developers," company CEO Robin Purohit told reporters. "And recently Clustrix has released a proprietary database 5.0 new version on the AWS platform. Of course, this version can also be run on platforms such as Rackspace, Gogrid, Equinix and Blueboxgroup. (in 2010, when the initial version of Clustrix was released, Gigaom reporter Stacey Higginbotham the product.) )

Clustrix Sierra

Clustrix Sierra is a distributed database system that supports both the simplest key-value format and MySQL functionality, and can be very simple to store billions of of data by adding new nodes. Its function is mainly embodied in the following four aspects:

Scalability and high availability: As long as you add one node, you can achieve linear elevation on performance metrics such as throughput, up to hundreds of nodes.

Speed aspect: It solves the situation that the traditional database writes the data slowly, and guaranteed the speed that reads the data.

Compatible with MySQL: It supports all MySQL features and supports acid (atomicity, consistency, isolation, durability) and is very handy for migrating existing MySQL.

On-line modification and extension: Not only can the new nodes be dynamically added in the production environment, but the table structure of the database can be modified online.

Operating mechanism

First, the Clustrix Sierra System consists of three core modules: one is distributed Query Compiler (distributed query compiler). The second is the distributed Shared-nothing execution engine (Execution Engine). The third is the storage engine (Storage Engine) that supports synchronous transactions (ransactional concurrent). And the whole system realizes its main function by integrating these three core modules.

Run the process

When many queries are sent to front-end receptacle, the system translates these queries into the system's internal format via the Database personality module, which is then passed through the distributed query mentioned above The compiler and execution engines run parallel queries, store data on the SSD through the storage engine, store log modifications on the NVRAM, and finally communicate and synchronize between multiple nodes through InfiniBand.