Disaster Recovery Solution

System disaster recovery is actually a big topic, including program disaster recovery and operation and maintenance disaster recovery.
Let's talk to you about disaster recovery in terms of operation and maintenance. When it comes to O&M disaster recovery, you may think of clusters. For example, nginx does application load balancing. This can indeed solve the problem that an application server node hangs without the application providing services. So what if the key node nginx hangs up. Then another solution is needed. Let's introduce the solution we are using now: LVS+Keeplived for high availability clustering and load.

Explanation of terms: LVS linux visual server is a linux virtual service.

The solution is: dual-machine hot backup architecture for key nodes, that is, two key nodes (nginx) mount multi-application web servers in the background. Between each server, keeplived is used to detect the status of other nodes. For example, two key nodes have master and backup nodes, namely active and standby nodes. If the master node hangs up, the backup node will detect it immediately, take over all the services of the master node, and at the same time modify its own IP to provide services. When the primary node is restored, the backup node will return the service to the primary node, restore it to the backup node, and release IP resources at the same time. . In this way, the active and standby dual-machine hot standby ensures that we can provide services stably. This is also a common disaster recovery solution in the industry. The

next section will talk about program disaster recovery.