A weekly technical update on distributed technology 2016.07.10

Distributed System Practice

1. Anatomy of the Elasticsearch cluster: storage model and read and write operations

https://mp.weixin.qq.com/s?__biz=MzA5NzkxMzg1Nw==&mid=2653159983&idx=1&sn= a79206bac0f14d61b47620ceed6cb0da&scene=0&key= 77421cf58af4a6535cd92e6d57c51e9de6a6de948c6455bdd581f7c300ae7af1b27f351f0c8f8313f3a21d79eeafb0cb&ascene=0 &uin=mjk1odmyntyymg%3d%3d&devicetype=imac+macbookpro11%2c4+osx+osx+10.11.5+build (15F34) &version= 11020201&pass_ticket=12oxbu0ts9tjolo2e1ugdxgjoi3gidolq57k3tspkx%2fv6ks7py%2belgidv1%2b%2fyczm

Key points: Elasticsearch because of its good business ecology, the position of the open source search engine is steadily climbing, especially the popularity of the micro-service architecture, to elk as the software stack to solve the service log analysis framework of the scheme is greatly respected, This article is a series of introduction Elasticsearch cluster's first article, introduced the Elasticsearch cluster's data storage model and the curd operation flow.

2. Message Queuing Design Essentials

Http://tech.meituan.com/mq-design.html

Important: This article is a very comprehensive description of the design of a message queue to consider the factors and various trade-offs, which I feel I have read about the message queue design of the best article, recommended to everyone.

Service and resource management technology

1. Research on Docker network scheme

https://mp.weixin.qq.com/s?__biz=MzA5OTAyNzQ2OA==&mid=2649690618&idx=1&sn= 03472edbd3d3038c72064bfa352b2f30&scene=0&key= 77421cf58af4a653126e86ffa400c9073e2b23bdfd253f0441782e35a97ab42c29c92e1e7b900cbfa76100606779821a&ascene=0 &uin=mjk1odmyntyymg%3d%3d&devicetype=imac+macbookpro11%2c4+osx+osx+10.11.5+build (15F34) &version= 11020201&pass_ticket=oozoqi574cl8jnjtaxrxc%2f8a4mxjaydae8vr3fs8%2bmvybyb5gy2gmnetg6j%2flgpd

Important: The Docker network scheme consists of two camps, one is the tunnel camp, the other is the routing faction, the tunnel camp needs packet and package, the performance cost is large, the routing camp needs to manage the complex routing table, the control overhead is large. This article describes in detail the calico that belong to the routing faction, including the fundamentals, demo demos and performance tests, and the dry goods are full.

2. Docker five storage Drive principle and application scenario and performance test comparison

https://mp.weixin.qq.com/s?__biz=MzA5OTAyNzQ2OA==&mid=2649690685&idx=1&sn= 9cbb9246f24f5d2b19aff8b92d9ea322&scene=0&key= 77421cf58af4a65376684a3b31af2daf8f5ceef6f7a348b23c4c94d9f7be7a188565c5601945fe260179c5579ed64fe4&ascene=0 &uin=mjk1odmyntyymg%3d%3d&devicetype=imac+macbookpro11%2c4+osx+osx+10.11.5+build (15F34) &version= 11020201&pass_ticket=nzjeordd%2f3upwnkpar03almxqgbtr5pv2njueteq1lmoenp2rkv2a3ou%2bfcotr9o

Important: Docker has 5 storage drivers, there is no perfect solution, this article describes the principles of 5 storage scenarios, and provides a performance test comparison, to help you follow the use of the scene selection.

Service High Availability Technology

1. The evolution of today's headline structure--The topic of architecture evolution under high pressure

https://mp.weixin.qq.com/s?__biz=MzAwMDU1MTE1OQ==&mid=2653547520&idx=1&sn= f303a6250eb68775e9b6dbbdea6b9f06&scene=0&key= 77421cf58af4a653fca7a21b63ea8260b9b2c62de529a260584803d0cefc60c8959acc44df470f45746589ab60ebb98a&ascene=0 &uin=mjk1odmyntyymg%3d%3d&devicetype=imac+macbookpro11%2c4+osx+osx+10.11.5+build (15F34) &version= 11020201&pass_ticket=ub1xoek8dvcbc4kf%2bkp9ixe4s2qedhdoexz4cfnclv%2bcc9aijj29%2ftxwprhx8u%2fs

Important: This article introduces the micro-service evolution process of today's headlines, and details the service autonomy in the process of evolution to microservices, and it seems to be one of the few domestic companies that use Consul to do service discovery.

2. The availability and performance optimization of a single point system architecture

https://mp.weixin.qq.com/s?__biz=MjM5ODYxMDA5OQ==&mid=2651959480&idx=1&sn= 337bd74410a6bef616128fd17abd08a8&scene=0&key= 77421cf58af4a653582894fea908eecacab1e9c68815b00ef9d2bdf906dd89364b1e0e459cc655d02b19e3066a5ea2ff&ascene=0 &uin=mjk1odmyntyymg%3d%3d&devicetype=imac+macbookpro11%2c4+osx+osx+10.11.5+build (15F34) &version= 11020201&pass_ticket=nzjeordd%2f3upwnkpar03almxqgbtr5pv2njueteq1lmoenp2rkv2a3ou%2bfcotr9o

Important: A single point system is often not completely avoidable (like the GFS master), this article describes several common scenarios for improving the availability of single-point systems.

DevOps Technology

1. Use big data thinking to do operation and maintenance monitoring

Http://www.jianshu.com/p/f634d7fc0f05

Key points: Now is the Big Data era, monitoring data is a very realistic and real big data application scenarios, including state data dashboard, trace call chain, business indicators, hardware failure, and so on, this article describes the big Data operation and maintenance of the idea of monitoring.

2. What features should be included in a complete microservices system?

https://mp.weixin.qq.com/s?__biz=MzA5Nzc4OTA1Mw==&mid=2659597647&idx=1&sn= 9a95fa4a12ae2cf6ce9a23ef9ad71400&scene=0&key= 77421cf58af4a65360d204600ef660ed60fd3185e57f66c414059151743e00a3c1cb6bbcc6ef879fde4c3a886afbd31f&ascene=0 &uin=mjk1odmyntyymg%3d%3d&devicetype=imac+macbookpro11%2c4+osx+osx+10.11.5+build (15F34) &version= 11020201&pass_ticket=ub1xoek8dvcbc4kf%2bkp9ixe4s2qedhdoexz4cfnclv%2bcc9aijj29%2ftxwprhx8u%2fs

Important: In recent years, the micro-service architecture is rapidly moving red in the entire technology community, it is considered to be the future direction of it software architecture, the great God Martin Fowler also give micro-service very high evaluation. So why do we need microservices, what is the real advantage of microservices, what is a complete microservices system, what features should be included, the author Liu Yanff has more than 10 years of experience in software design and development, and he will answer from his perspective.

Hodgepodge

1. What should I do with performance testing?

http://coolshell.cn/articles/17381.html#rd?sukey= 3997c0719f151520652bc226bcac03214a8ea5fabeae2b65ae70a171436d00dd4e6130edfc2804375f9229045c3806e0

Point: We do the architecture of the engineers must be inseparable from the performance test, then your performance test plan is correct? This article tells you how to conduct a performance test scientifically.

2. How do technical teams discover and nurture tech leads?

http://mp.weixin.qq.com/s?__biz=MjM5MDE0Mjc4MA==&mid=2650993234&idx=2&sn= 8d1d30846352624e127fa63a50d4693a&scene=1&srcid=0706cjksxly3z9oysyrovz0r&from=groupmessage& Isappinstalled=0#wechat_redirect

Key points: The so-called thousand armies are easy to get, one will be difficult to find and cultivate tech lead is not a simple thing, whether it is a new classmate or old driver, this article will have a harvest for you.

A weekly technical update on distributed technology 2016.07.10