Internet Architecture Design (high concurrency high availability) _ High concurrency

The 1th part of the architecture is summarized. 第2-8部 the technology used at each end, no details. Parts 9th and 10 are used to monitor and deploy, troubleshoot, and resolve problems.

One: Overview of server Overall architecture

The stage of experience:

1: Application service and Data service, application dedicated picture server, video server.

2: Cache improves website performance, redis,memcached cache.

3: Application server cluster to improve the concurrent processing capacity of the site, this time should be read-write separation or dual-master read and write separation.

4: Use reverse proxy and CDN Acceleration website

5: Distributed database system, database partitioning stage

6:nosql servers to ease the hassle of application management of many data sources.

7: Business split, split a site into many different applications, each application independently deployed.

After the above seven stages, the overall server architecture is as follows:

Second: Database technology

Separation of application services and data services;

Database read-Write separation

Distributed Database System
Sub-Library Table

Database index

Database Connection Pool
SQL optimization

Database-Related Settings

Three: Caching technology

Use caching to improve site performance;

Increase cache Hit rate;

Redis cluster related settings;

Four: Back-end technology

redirect

Use reverse Proxy

DNS polling, unable to assign tasks by server capability
CDN, and according to the flowmeter fee, the price is also more expensive.
IP load Balancing: F5,vs/nat (based on network address translation technology), Vs/tun (ip-based tunneling Technology) and VS/DR (based on direct routing technology)

Mirroring (the education industry in which students are in use)

V: Business Split

Split a Web site into many different applications, each of which is deployed independently. Applications can be established through a hyperlink (each of the navigation links on the home page points to a different application address), or through Message Queuing for data distribution, or, of course, the whole system that makes up a connection by accessing the same data storage systems.

VI: Front END

Page static
Front-End caching
IOS, Android database use

IOS, Android self file use

Eight: Other

Replace Apache as Nginx;
The queuing system is on the scene, taking RABBITMQ as an example

MySQL Hardware settings

Nine: monitoring software

Najios Monitor service running status and network information;
Zabbix Monitor service running status and network information;
Cactic the machine information and data information graphically to the user.

Ten: Deploying Software

Saltstack
Chef
Puppet
Ansibel