Server Redundancy

Server redundancy means that, if a primary server is unavailable due to a power outage, network connection loss, or other failure, a designated backup takes over and enables the application to continue to run.  This applies to both configuration servers and service servers.

Backup servers behave as clients until their designated primary fails. If the primary server fails, the first workstation to be designated as a backup server immediately takes over the primary’s role, enabling the application to continue running until the primary server is back online. The following example shows how this process works.

Given 3 workstations:

•    Primary configuration server A,

•    Backup server B, and

•    Client C.

While A is functioning, both B and C operate as clients of A. This relationship is illustrated in the image below.

If A is shut down, B immediately assumes the role of server, and enables the application to continue to run. C is now a client of B. While B is operating as backup server, its clients will only be able to complete operational changes (see Operational Changes vs. Configuration Changes). Configuration changes can only be made when the primary configuration server is online.

When the primary configuration server restarts, it synchronizes all application and data files with the backup server, and then reassumes its role as primary configuration server, while the backup server resumes its role as a client of the primary configuration server. In our example, as A restarts and the VTS application begins to run, A looks for a server so that it can get a copy of the most up-to-date data. A is now acting like a client to B, and synchronizes its application data and files with those of server B. This is illustrated in the image displayed below.

Once synchronized with B, A's internal synchronization mode instantly drops to 0, and A and B reverse roles; A assumes the role of primary configuration server, and B is once again a client to A.

In the event that client C is communicating with B while B is acting as the backup server, and a switch of servers occurs (i.e. the primary configuration server comes back online), client C switches to A and synchronizes its application data and files with those on A. In other words, each time a server changes, all clients resynchronize their application data and files with that of the server.

If the primary configuration server and the first backup server fail, the second workstation to be designated as a backup server assumes the role of server.

 

The process used by VTS to provide server redundancy is illustrated by the following diagram: