Correlation IDs bundle one transaction as it moves through multiple processors. With this system, your client's requests are collected under one header for easier tracking and troubleshooting.
What is correlation ID?
We often think of a visit as a one-and-done transaction. In reality, one visit could spark several different processes. Correlation ID keeps all of the work together.
Let's start with an example.
You run an auction website that provides items from hundreds of sellers. Your visitor buys from three sellers in one order. Your website splits those orders to the disparate sellers. A week later, your client asks where the order is. A correlation ID could help you track all of the pieces that make up this one order.
Typically, a correlation ID is a nonstandard HTTP header, and it's part of the Java Messaging Service (JMS). You don't have to use one, but it's wise if you have even the smallest opportunity of using multiple processors to handle client requests.
5 ways to use correlation ID
Most programmers can set up correlation IDs in minutes. If you're adept at code, figuring out the logistics can be a little tricky. But here's how it works in a simplistic form.
Here's how to use correlation ID:
- Create. Developers put a randomly generated identifier in every request, typically as part of the HTTP header.
- Assign. Your IDs stay with the order as it moves through sub-processors.
- Log. Keep a record of each correlation ID you create, so you can handle client requests as they appear.
- Transport. How you move correlation IDs to other systems (including sub-processors) will depend on how your system is built and maintained.
Should you use correlation IDs?
If you run a simple business with easy processing, you may not need correlation IDs. But if your workflows include sub-processors, assigning these unique numbers could help you provide optimal customer service.