Stop hunting for resource leaks in systemd services
Process & Resource Monitoring (Top/Htop/Ps/Systemd)
Stop hunting for resource leaks in systemd services
🧩 The Challenge
You know the feeling when a service starts eating RAM at 3 AM and the memory usage doesn’t show up in standard process lists because it’s managed by systemd? I’ve spent whole shifts guessing why a database exporter kept crashing only to realize systemd was throttling it based on slice limits I totally forgot I set.
💡 The Fix
Use the systemd-cgtop utility to get a live, top-like view of your control groups so you can see exactly which service is burning through CPU or memory cycles. It cuts through the noise and shows you the resource hierarchy instead of just a pile of PIDs.
systemd-cgtop
⚙️ Why It Works
This tool taps directly into the cgroup v2 controller, which is the ground truth for how systemd allocates hardware resources. Because it groups processes by their unit name, you’ll spot a runaway service in seconds rather than scrolling through hundreds of lines in top.
🚀 Pro-Tip: If you need to see exactly which slice is hitting a hard limit, hit ‘o’ to toggle the sorting and look for the services pegged at 100 percent of their allocated share.
Linux Tips & Tricks | © ngelinux.com | 10/30/2026
