Strange CPU and RAM Usage Increase in Python Docker on AWS – Help!

So far, when I run the relevant parts of the code locally, I've not encountered any problems. The memory usage seems stable for me.

Double-check your I/O operations too! Sometimes, if an I/O reader isn't properly closed, it can lead to resource leaks. For instance, if you do something like `f = response.Body.Read()` but forget to follow it with `response.Body.Close()`, that could cause issues.

Have you tried reproducing the issue locally? The fact that it's running on AWS shouldn't change anything significantly. It might be a problem with your code not managing memory properly. For example, check if you have any variables that are being reused instead of overridden, which could lead to memory build-up.

This seems like a specific application issue. Just calling gc.collect() won't fix reference cycles. Make sure you're not holding onto objects in ways that prevent them from being garbage collected. Enable debugging for the garbage collector with gc.set_debug(gc.DEBUG_LEAK) to track object creation. Also, look out for unreleased references, global variables, and any open file or socket handlers that could be causing leaks.

Using a memory profiler could be a great next step. Python has a built-in tool called [tracemalloc](https://docs.python.org/3/library/tracemalloc.html#compute-differences) that can help you monitor memory usage over time. While memory_profiler is another option, it seems it's no longer maintained, so keep an eye out for alternatives. With these tools, you might identify growing structures or leaks in your code.