Server redundancy provides automatic synchronization of Point and Alarm Databases. Server redundancy provides automatic switchover of CimView application using Points and Alarms. Before you start building your application you should review the section in this documentation entitled "Limitations of Server Redundancy", to verify that the features of CIMPLICITY that you intend on using are supported in server redundancy.:.
CIMPLICITY will run your application as you design it. CIMPLICITY cannot automatically fix your project if you design it incorrectly. Therefore, it is important that you design your project to be mission critical from the ground up. Also, it is imperative that you test your application in a server redundant environment with viewers during the development stage. Only with a properly configured project can you switch on server redundancy and have it work flawlessly.
We at GE Fanuc have designed many redundant systems using CIMPLICITY. We understand the methodology and design techniques needed to build a robust system. Therefore, we do recommend contacting your salesperson to obtain several days of design consultation before you start your first project, and several days of on-site support during deployment.
Scripting Requirements for Server Redundancy
The single biggest issue in building a server redundancy system is your user defined scripts. During failover, point values may be unavailable for a short time. Scripts must be written to properly handle these intermittent periods and to exit cleanly. Scripts that depend on cleanly exiting must be coded to trap the errors that can occur when a point goes unavailable. You must test your scripts during fail over to verify they operate correctly.
Use of Primary / Secondary Computers
The purpose of your primary and secondary computers is to read and process data from your devices, distribute it to viewers, and to remain synchronized. They need available CPU bandwidth to handle exception conditions in your process. If you have viewers in your system, the primary and secondary servers should not run user interface applications such as CimView. The secondary server is not a "spare" computer to be used to perform other chores like word processing, etc. It is a hot backup, dedicated to providing redundancy for your mission critical application.
Important: The primary computer must have a mapped drive to the secondary computer. It is through this mapped drive that a qualified user (a user with administrative privileges) can start and stop the slave.
Database Logging Requirements
If you are planning on using database logging, you should certainly read the information in this document on how to use logging within a server redundancy project. Additionally, in a mission critical application, the use of Microsoft Access as a database is not supported. Instead, Microsoft SQL Server, Oracle, or other supported database server must be used. If you plan on logging a large volume of data you may want to consider locating the database servers on separate computers within the same LAN / switch as the primary and secondary. Remember the total CPU utilization, including the database server, must be less than 40%.
Network Configuration Requirements
In addition to having a solid physical network, server redundancy requires specific network software configuration to be performed on every computer in the system. Since specific configuration is required on every computer you cannot just "plug" another viewer into the network and expect it to work. The network configuration must be updated on the viewer and related computers.
Time Synchronization Requirements
The times on the Primary and Secondary computer must be synchronized. Additionally, if using trending on viewer computers, the times on the viewers must be synchronized with the servers. It is your responsibility to ensure that the computer times are synchronized. There are a variety of commercial products available to maintain time synchronization between computers. If you choose to automatically synchronize your clocks do so at any time other than midnight.
Server redundancy overview.