The documentation talks about adding RDI Server instances. Why would you want to do this? Are there advantages?
Depending on the number of tags and RDI’s I would definitely add more RDI Server instances to large systems. Here are some of the advantages:
- It helps while troubleshooting issues with the RDI Server such as narrowing down which RDI has problems that lead to memory leaks
- It allows you to connect to multiple data sources that use different usernames and passwords
- It lightens the processing and memory load for a single RDI Server which allows you to more efficiently use a server’s resources
- With multiple small RDI Servers, stopping and starting PHD is a faster process
If you have one server with one data source and less than 10k tags then it is not needed. If you have more than one data source I would build an additional instance for each one. I would also limit the number of tags per RDI server to around 40k max.