What Linux tools or workflows genuinely changed how you think about system performance and resource usage?

I've been spending a lot of time lately digging into how Linux exposes system internals in ways that Windows and macOS simply do not. Being able to inspect scheduler behavior, tune swappiness, or use tools like perf and htop to get a real picture of what is happening under the hood has completely shifted how I approach performance problems.

It started small, learning that not all I/O schedulers behave the same way depending on whether you are on NVMe or spinning rust. Then I fell down the rabbit hole of cgroups, CPU affinity, and realizing that the defaults on most distros are reasonable but not always optimal for specific workloads.

What I am curious about is whether others have had similar moments where a specific tool or concept fundamentally changed how you approach your system. Not just a handy command you learned late, but something that genuinely reframed how you think about computing resources.

For me it was combining perf stat with flamegraphs to actually see where CPU cycles were going in a longrunning process. Before that I was basically guessing.

Are there particular areas like memory management, networking, or storage where you found Linux gave you insight you simply could not get on other platforms? Would love to hear what tools or subsystems clicked for people and why.

https://redd.it/1ugw4p6
@r_linux