Proxy and Parallel Servers
Proxy and parallel servers add efficiency to large configurations, and can limit the damage caused by an outage.
Both configurations are a type of distributed configuration, but they differ in how the workload is distributed between the servers:
- In a proxy environment, each proxy server takes a portion of the workload and works on that portion exclusively.
- In a parallel environment, the workload is distributed among all of the servers, with one server attempting to keep the distribution as balanced as possible.
Proxy servers are particularly useful for systems that are widely dispersed over a large geographic area. In a proxy environment, the number of clients attached to a server is distributed across a set of servers (running in proxy mode), all of which funnel down to a central server (the Master).
Parallel servers enable load sharing. That is, multiple instances of a server run in parallel, and the load is distributed among them.
The use of proxy and parallel servers greatly reduces the loss of functionality and data if a server goes out of service.
- If a proxy server fails, you lose only the clients associated with that proxy server. In a non-proxy environment with only one server instance, if that single server goes down, all the clients are lost.
- If a server in a parallel configuration fails, new requests are distributed to the remaining servers.
Refer to documentation for your product to determine if it supports some variation of proxy and/or parallel configuration, redundancy, and how to implement it for your system.
Proxy and parallel servers address the efficiency issue inherent in large configurations, and also minimize the loss of functionality and data in the event of an application failure.
Proxy servers are particularly useful for systems that are widely dispersed over a large geographic area. In a proxy environment, the clients that require a connections to a server are distributed across a set of servers (running in proxy mode), all of which down to a central server (the Master). If one proxy server fails, only the clients connected to that server are lost. Compare this to a single server environment, where, if the single server fails, the whole system is lost.
Parallel servers enable load sharing. That is, multiple instances of a server run in parallel and the load is distributed among them. If one of the parallel servers fails, new requests are distributed to the remaining servers so there should be no loss of service.
Each server in a Proxy or Parallel configuration can also be set up with a backup server, enabling each to take advantage of the benefits of redundancy. Refer to Application Redundancy.
The configuration of proxy or parallel Genesys servers can vary, depending on the server type. For more information, and for detailed instructions for setting up proxy or parallel servers in your environment, refer to the appropriate product documentation.