One common addition to timestamps is correlation IDs. Basically, they're unique identifiers that get assigned to each traced message. Their purpose is to correlate the logs produced by different components of your system (like different microservices) during the processing of the same incoming request and sometimes for the events it caused, too. Such IDs should be passed with the message everywhere it goes, for example, by appending to its HTTP header. Even when the original request is gone, you could add its correlation ID to each of the responses produced.
By using correlation IDs, you can track how messages for a given request propagate through the system and how long it took for different parts of your system to process it. Often you'll want additional data to be gathered along the way, like the thread that was used to perform the computation, the type, and count of responses produced for a given request, or the names of the machines it went through.
Tools...
