Difference between revisions of "Load balancer"
| Line 1: | Line 1: | ||
| − | A [[load balancer]] (hereinafter, the ''Balancer'') is any [[system]] that performs [[load balancing]]. The ''Balancer'' accepts client's requests, selects | + | A [[load balancer]] (hereinafter, the ''Balancer'') is any [[system]] that performs [[load balancing]]. The ''Balancer'' accepts client's requests, selects one resource among two or more to fulfill that request, forwards the request to the selected resource, receives its response, and returns that response back to the client. [[Computing node]]s, [[DBMS]]es, and [[application]]s may serve as the resources as long as they can be found at particular [[IP address]]es. |
The ''Balancers'' work with multiple nodes; this feature distinguishes them from [[reverse proxy|reverse proxi]]es that accepts client's request, forwards this request to the resource, received its response, and it to a server that can fulfill it, and returns that response to the client. | The ''Balancers'' work with multiple nodes; this feature distinguishes them from [[reverse proxy|reverse proxi]]es that accepts client's request, forwards this request to the resource, received its response, and it to a server that can fulfill it, and returns that response to the client. | ||
| + | In addition to their core purposes, the ''Balancers'' often detect failures of the resources using daemons such as [[heartbeat]], [[pacemaker]], or [[keepalived]]. | ||
| + | |||
| + | |||
| + | ==COTS== | ||
| + | *[[IP Virtual Server]], that performs transport-layer load balancing inside the Linux kernel, so called Layer-4 switching. | ||
| + | *[[HAProxy]]. | ||
| + | *[[NGINX]]. | ||
==See also== | ==See also== | ||
Revision as of 20:02, 9 June 2021
A load balancer (hereinafter, the Balancer) is any system that performs load balancing. The Balancer accepts client's requests, selects one resource among two or more to fulfill that request, forwards the request to the selected resource, receives its response, and returns that response back to the client. Computing nodes, DBMSes, and applications may serve as the resources as long as they can be found at particular IP addresses.
The Balancers work with multiple nodes; this feature distinguishes them from reverse proxies that accepts client's request, forwards this request to the resource, received its response, and it to a server that can fulfill it, and returns that response to the client.
In addition to their core purposes, the Balancers often detect failures of the resources using daemons such as heartbeat, pacemaker, or keepalived.
COTS
- IP Virtual Server, that performs transport-layer load balancing inside the Linux kernel, so called Layer-4 switching.
- HAProxy.
- NGINX.
