I have a server application. I log each request when it begins and when it ends (or errors).
Additionally, I create a unique request id
when they are received and add it to its logs, and also add a timestamp
for begin and end.
If another child request is created, I do the same, and also include the parent's id
. I add to the log some useful data like request method
, user id
, etc. as it becomes available in each child request.
Is this kind of logging exactly the same as tracing? In other words, if I'm doing this, there's no need to do tracing, since it's exactly what I'm already doing?