Open-source nosql system membase

Membase is a new heavyweight Member of the nosql family.

Membase
Steps for installing and using a cache server

See: http://www.oschina.net/question/12_12770

Membase is an open-source project with the source code license of apache2.0. This project is hosted on GitHub. Source
Currently, you can download the beta version of the Linux Binary Package. This product is mainly developed by members of the memcached core team of North scale. It also includes engineers from Zynga and NHN, two major contributors, both organizations are large suppliers of online games and community network spaces.

Membase is easy to install and operate, and can be easily expanded from a single node to a cluster. It also implements the plug-and-play function for memcached (compatibility of Wired Protocols) and is available for applications.
The sender and operator provide a relatively low threshold. As a cache solution, memcached has been widely used in different fields (especially for large-capacity Web applications). Some of the Basic Code of memcached has been directly applied to the front-end of the membase server.

Membase has good reusability by compatibility with multiple programming languages and frameworks. In terms of installation and configuration, membase provides an effective graphical interface and programming interface, including configurable alarm information.

Membase aims to provide external linear scalability, including the ability to replicate unified nodes to increase cluster capacity. In addition, redistribution of stored data is still necessary.

An interesting feature in this regard is the predictable performance promised by the nosql solution, latency and throughput of class accuracy. The features mentioned above can be obtained through the following method:

• Technologies that automatically migrate online data to low-latency storage media (memory, solid state disks, disks)
• Optional write operations: asynchronous and synchronous (based on replication and persistence)
• Reverse Channel rebalancing [support for future considerations]
• Multi-thread low lock contention
• Use asynchronous processing whenever possible
• Automatic deduplication
• Dynamically rebalance existing clusters
• By copying data to Multiple Cluster units and supporting fast failover, the system provides high availability.