Co-authors of this article: Wang Pengcheng (Director of Operation and Maintenance); Jia Xiaohui (Head of Cloud Company); Han Xiaoguang (Head of Operation and Maintenance)
20 rules of operation and maintenance automation summarized by community experts
1
Original intention : The operation and maintenance automation system should be practical and able to support the business well, not just a pure technical theoretical product.
2
A good operation and maintenance system architecture : reliable and stable, simple and practical, and suitable for business needs .
3
The core architecture design is not good : waiting for infinite rework later, or simply stop the work.
4
Design 28 principles : The operation and maintenance system architecture design is not easy to complete, the functions should be refined and specialized, the modules should be decoupled, and it should not be over-designed.
5
It is not necessary to automate operation and maintenance to be comprehensive and comprehensive. There is no need to automate operation and maintenance to the extreme of technology. It is a good choice to fit the business and be small and precise.
6
Life cycle principle : The object handled by the automated operation and maintenance system is the full life cycle of the CMDB, which is a full-process closed-loop system integrating collection, monitoring, management, control, disposal, and public opinion .
7
Development tools : Operation and maintenance personnel recommend choosing python development and operation products. Bootstrap can be used for the front end, and django and flask can be used for the WEB framework. For cool animations, vue and Grafana can be selected. Conventional data is stored in mysql, performance data, and trend analysis. Data can be selected from time series databases, such as infludb, RRDTool, etc., message cache using redis, Kafka, historical data can be placed in hbase or mysql. Many other excellent software can be used.
8
Advanced development : Make good use of the following keywords to make you do more with less, paramiko (ssh network communication module), Celery (distributed task scheduling), epoll (asynchronous multi-threading), threading (multi-threading module), queue (queue module) , twisted (based on event network module) supervisor (non-daemon process management), subprocess (system interaction module), urllib2 (network access), CBV development mode.
9
What kind of team is more suitable for developing an operation and maintenance system : the design and implementation of an operation and maintenance automation system is best to choose a team that understands both operation and maintenance, and a team that only understands development or only understands operation and maintenance. automated system.
10
Operation and maintenance system goals : To implement a business-oriented operation and maintenance automation system, the team needs to understand both development and operation and maintenance, as well as business.
11
Standardization : Design operation and maintenance automation, CMDB system, first of all, you need to standardize, determine the resource CI, sort out the resource relationship, organizational structure, technical structure and process specification.
12
Pay attention to data processing : determine the monitoring indicators and information collection methods to ensure the consistency and correctness of data collection and changes.
13
Principles of software engineering: The operation and maintenance automation software system is software engineering, which follows the rules of project engineering.
14
Library table design : The database table structure design is related to what data you want to collect and display, and affects the complexity of your program logic.
15
Database Design : Database choices affect your storage architecture, distributed architecture, and performance issues.
16
Tool selection : Select product tools with active communities and high ratings, not necessarily the latest.
17
Implementation methodology : Don't easily repeat the creation of wheels, and check whether there are excellent ideas and products in the industry before making them.
18
Feature Highlights : Data connection between product modules, streaming and multiplexing are both difficult and successful highlights.
19
The security of operation and maintenance automation is a double-edged sword : special attention must be paid to security protection and authority control, some high-risk actions must be avoided, and rm * execution must be avoided, and a sound operation and maintenance security and audit system must be established.
20
Opportunities and risks are the same : Operation and maintenance automation brings dawn and opportunity to operation and maintenance personnel, but don't expect a set of operation and maintenance automation system to free the operation and maintenance personnel from all work. Humans have planes, but they can't do without trains. Humans have elevators, but they can't do without stairs. Why?
Want to meet more automation operation and maintenance experts face to face?
Come to DevOps International Summit Beijing Station-Automated Operation and Maintenance Session
Four experts from Ali, JD.com, Huawei, and China Merchants Bank unlock more new skills for automated operation and maintenance for you
The preview video of the conference is freshly released⬇️
Click to read the original text , more exciting