[Answered ]-How to understand the profile output of a Django view function

Your data descibes how many times a given function was called and how much time was spent within that function.

• ncalls is the number of times that the function was called
• tottime is the total amount of time spent within that function, excluding subfunctions
• percall is the average amount of time that function took per call, or tottime/ncalls
• cumtime is the total amount of time spent within that function, including subfunctions
• percall is similar to the first percall but now for cumtime

So this tells you that a large amount of time is being spent in socket.py. Specifically, readline is called a few times and takes a pretty large amount of time each time. create_connection is called only once but that takes even more time than a call to readline. And within posixpath.py, islink is called a huge amount of times but takes a very small amount of time during each

Given that create_connection is only called once(which makes sense), there’s likely not much you can do to change that.

It’s possible that you could reduce the number of times that readline is called by figuring out when the function is called and changing your code to avoid multiple calls to readline when one would suffice, but I don’t have enough knowledge of your code to say whether that’d be doable.

It seems very likely that you could reduce your calls to islink but it would only make a small difference in speed.

Really the only way that you’re going to dramatically improve the speed of your code is by reducing those readline calls so I’d focus on figuring out when it gets called and minimizing those calls.