The road of DB cluster one

Recently, the front-end angular development framework has been set up, and the framework, tripartite Library, jquery and jquery plug-in library have been successfully solved. Within the framework of the implementation of the control package, such as pagination control, layout control, selector control, drop-down selection control, and then to the form life cycle, dynamic creation of components, page routing, module collation, routing Guard, HTTP interception, SSO, file upload, scene animation and so on each node realization, Basically fully meet the existing requirements of functional development, I think back to the front end is the chart, Gis,websocket a class of applications combined.

On the back end, for a distributed management system is the SSO unified authorization, WEBAPI can, through the SSO life cycle analysis, the establishment of the session model, and the authority mechanism, and then to WEBAPI infrastructure, routing, message filtering and unification, HTTP context management, and so on, Basically it can be said that the whole management platform of the basic key nodes are well solved and implemented.

For a distributed system to just let all the API to access a database is certainly not possible, so long ago to understand the database master-slave architecture, his first step is to read and write separation, well, online to see a variety of posts one of the very good MySQL master-slave architecture analysis of the topic

Http://www.cnblogs.com/crazylqy/category/625963.html bloggers are very well written and very easy to understand when they are posted. Later found that these are not enough to meet the distributed database architecture, then encountered a new thing called mysql-proxy (for me new), my dear originally this is a database cluster of an excessive, with the concept of the agent, the original agent to solve the problem of read and write separate SQL routing, The presence of agents constantly solves SQL routing, filtering, load balancing, and finally achieves a high-availability database clustering architecture. Originally in this field this agent is actually a middleware tool, such tool software has a lot of open source and many such as mysql-proxy,amoeba,haproxy and so on, but Baidu Baidu do not understand these things are just say how to build a cluster, what level of split, Vertical splitting and so on I'm so dizzy, I just want to know how to use this middleware for client ORM, how to use Ah, but people do not pay attention to this matter, well, it is best to find the domestic use of high frequency and open source, and still have someone to maintain the state is mycat, a domestic open source and technical comparison cattle , the database cluster open-source middleware tools, while providing net,java,php a class of platform language development interface, it seems to take some time to learn the next mycat, after all, distributed system first at the database level should be distributed, and then solve the application server distributed.

The next stop is Mycat, and now there is a question about how the existing ORM can be interfaced with the results provided by MYCAT.

The road of DB cluster one