The data-collection system architecture should support not only the functional specifications listed earlier, but also the nonfunctional specifications, such as high-availability, fault-tolerant, distributed, and reliable system.

This section will present you with two sets of opinionated high-level architecture for the data-collection System.

The first one is based on service-to-service interactions and the second one diverges a little and makes the communication message-driven. The architecture for the second one will build on top of the architecture for the first one.

Both the architectures will follow the principles of micro services. We will develop the data-collection application as a suite of individual services with a specific responsibility. At a very high level, it will look like this:

I have a very specific reason to bring up such a picture right now.

I wanted to talk about the two main components depicted in the preceding figure:

• Service gateway
• Discovery server