Enterprise Distributed Logging – Necessary Constructs!

Service Bus for Windows Server is my friend, why? you might ask. I have been closely tied to this friend for over a year as I created a large distributed logging framework for a Client. Once of the realizations that I had is that in Enterprise Integration where so many different systems must interact; it is difficult to correctly aggregate all of the data and satisfy both Business and Technical teams. 

What I was able to induce from my time with the Service Bus , building out a complex logging ecosystem are a set of constructs that all should adhere to.

Here they are:

Logging is not just for errors.
Synchronous vs Asynchronous
No retry logic implemented
No fall back strategy
Single Point of Failure
No Data Persistence and Replay Capabilities
No Logging level specified
No correlation identifiers
No Priority set on the logs