In this article, we are going to talk to you about what is Load Balancing and how does works Load Balancer?
To understand the concept of Load Balancing, we must first understand the meaning of its words. Load is the number of resources consumed by the host, server, or host system, which can be hardware resources such as CPU and server RAM or network resources such as bandwidth and software resources such as servers. Each of these cases has a certain power that excessive requests from this power cause in response to requests that the amount of pressure on these resources is determined by the concept of Load. Currently in this case balancing means running the server resource consumption by managing requests can specify such cases that for the concept of Load Balance requires at least two separate servers in the case of another service called Load. The balancer can manage requests that are sent to the server.
The program described above Load Balancing is performed using a service with a Load Balancer identity and at least two servers. The image above shows a traditional hosting that is hosted directly on a single server, and in the image below you can see a schematic of the Load Balancing service. Similar requests to open a Web application or view a site by users are first made by Load Balancer and then each of these requests is sent to them according to the load and the number of servers, which causes this. That even in cases where a lot of requests are sent to an application site or can be distributed on the host servers and will be slow and we will not see the servers. The future of hosting servers is moving in this direction and now they can have this service in cloud servers, which Server.ir collection has provided the possibility of using this service in its cloud servers called aero server.
But what services are not Load Balancing? You may be familiar with the concept that in the past with traditional services it was possible to set up your site or service on two separate servers and then connect the domain to these servers through NameServers Kurds. But there are drawbacks to this method that make it different from the concept of Load Balancing. The first point is that in this type of implementation, only if the first server is out of reach, requests will be sent to the second server, and in fact, no balance will be done on the load servers, and all requests will be sent to the same server until the first server is disconnected. And then the second server will receive the requests, which in fact is equivalent to the concept of Redundancy. The second point is that the definition of NameServers in the DNS of the domain you are using only uses the Network access factor to ensure the availability of the server, which means that if a service similar to Web Server is unavailable on the host server, your site will not be available. Because the ping and DNS of the host server are available, the request will not be sent to the second server, and as a result, even though the second server is ready to make requests, it will not receive a request and you will see your site or service terminated.