The typical PI System consists of numerous components, with PI Server components installed on several server machines. While historically the PI system was installed on physical servers, installation on non-physical servers is becoming increasingly common.
Two main options are available for non-physical implementation:
- implementing virtual servers on a physical host that the business controls
- implementing servers on a cloud host controlled by a third party
This white paper addresses best practices for implementation of the PI System on the Amazon Web Services cloud service. Areas of focus include, but are not limited to: PI server configuration, network reliability and level of service, securitypractices, IP addressing, resource allocation, and disaster recovery.