Container Basics On Linux (Namespaces/Cgroups)
Peek under the hood of your container cgroups without the headache
🧩 The Challenge
Trying to figure out why a pod keeps hitting a memory limit is a nightmare when you’re just looking at a wall of YAML. I’ve spent way too many nights guessing which cgroup file the kernel is actually enforcing.
đź’ˇ The Fix
Skip the high-level orchestration tools and look directly at the sysfs entries the kernel manages. It’s the only way to see the raw numbers the OOM killer is looking at.
findmnt -no TARGET | grep cgroup
cat /sys/fs/cgroup/memory/system.slice/docker-<container_id>.scope/memory.usage_in_bytes⚙️ Why It Works
By navigating directly to the cgroup hierarchy, you bypass the abstraction layers that usually hide the actual resource accounting. Everything in Linux is a file, and the kernel exposes these performance metrics exactly where you’d expect.
🚀 Pro-Tip: Pipe that output into numfmt –to=iec-i if you’re tired of counting trailing zeros on those byte values.
Linux Tips & Tricks | © ngelinux.com | 10/17/2026