Large Web site technology architecture-core principles and case studies-read Note 5

In the fourth chapter of the case section of the mass distributed storage System Doris of the high-availability architecture design analysis, the author mainly analyzes the high availability of distributed storage architecture and high-availability solutions in different fault situations two aspects, in two subsections before the author introduced to us the Doris is a mass distributed kv storage system, It is designed to support medium-sized, highly available, scalable kv storage groups. Compared with the mainstream NoSQL system HBase, Doris has similar performance and linear scaling capabilities, with better usability and a more user-friendly GUI. In the section of the high-availability architecture of distributed storage, the author analyzes the overall architecture of Doris, which can be divided into application server, data storage server, and central Administration server in three aspects, in which the data storage server has set the number of copies of data according to the practical availability level. That is, a copy of the actual physical storage of each data tree, the more copies score, the higher the availability level, of course, the more servers you need. In the high-availability Solutions section of the different failure scenarios, the author has explained that the fault classification of distributed storage systems, the normal system access structure, the high-availability solution for transient failures, the high-availability solution for temporary failures, and the high-availability solution for permanent failures, five, in which the authors enumerate the instantaneous, Temporary, permanent three aspects of the failure, and in the following three subsections, the author specifically describes the three failure of the high-availability solution.

In the fourth chapter of the case of the website of the second Kill system architecture case analysis, the author tells the technical challenges of the second kill, the second-kill system and the second-kill system architecture design three aspects, including the technical challenges to the existing website business impact, high concurrency applications, database load; Sudden increase in network and server broadband , direct orders four aspects of the content, the strategy also contains a separate deployment, page static network bandwidth dynamic generation of random next page URL four aspects. The architecture of the second-kill system describes how to control the seconds to kill the product page of the purchase button and how to allow the first submitted order to be sent to the order subsystem two aspects of the elaboration, through reading to understand that the second kill is a great test of the site, again difficult to anticipate and control the impact of high concurrent access, A slight carelessness may cause a big problem.

In the fourth chapter of the case of the large-scale web site typical failure case analysis, the author talks about writing logs, high-concurrency access to the database, high concurrency of the index, caching, application start out of sync, large file read and write, misuse of production environment, irregular processes, bad programming habits nine aspects of the fault caused by In the different fault situations, the author has carried on the fault phenomenon, the reason and the experience lesson is that aspect concrete elaboration. When we write, we must pay attention to their programming norms, develop good habits, in the discovery of errors to learn a fall into your wit can not always make the same mistake.

Large Web site technology architecture-core principles and case studies-read Note 5