Displaying present location in the site.
August 25th, 2021
Machine translation is used partially for this article. See the Japanese version for the original article.
This time, as an introduction, we will explain the basic words of the HA cluster system.
Let's work hard together to become HA Cluster Master.
Now, let's get started.
Cluster Systems and Nodes
A cluster system is a system in which multiple computers are connected by LAN, etc., and behave like a single system.
A node is the server that is configuring a cluster system, and a cluster system is a group of two or more nodes that form a single system.
Could you have a little images that a cluster system consists of multiple nodes and works as a single system?
There are several types of cluster systems.
- Load Balancing Clusters : Clusters for workload balancing
- HPC(High Performance Computing) Clusters : Clusters for improving business performance
- High Availability(HA) Clusters : Clusters for business continuity
"HA" stands for "High Availability" and it is said that the ability to continuously operate business even if failure occurs on a server. In other words, an HA cluster is a cluster system for improving business availability that uses multiple servers and makes them redundant to minimize system downtime even if failure occurs.
For example, if a server running a business application fails, take over the business application to another server. By letting another server take over business applications, we can keep business ongoing.
Active Server and Standby Server
An HA cluster consists of multiple servers, but the name depends on the role of the server. The server on which the business application is running in the HA cluster system is called an active server. On the other hand, a spare server for switching business applications, such as in the event of a failure, is called a standby server.
As shown in the figure above, it is called failover to for the standby server to take over business applications on the active server by detecting a network, server, or other failure.
The standby server takes over the business application, so that the business application can continue to run. One of the points of failover is to stop the failed business application on the active server once instead of leaving it running, and then a standby server will take over the business application. In this way, remember to stop the business application once during failover.
This time, as an introduction, we explained about Cluster Systems, Nodes, HA clusters, Active server and Standby server, and Failover. In the next part, we will explain HA cluster configurations and HA cluster topologies.