Stop watching your load average lie to you about CPU spikes
Process & Resource Monitoring (Top/Htop/Ps/Systemd)
Stop watching your load average lie to you about CPU spikes
🧩 The Challenge
Everyone has seen that moment where the load average is sky high but the system feels perfectly responsive, or worse, the server is crawling and the load numbers look totally fine. I spent half a morning once chasing a phantom process that was just pinning one core while the others slept, completely masked by the load average aggregation.
💡 The Fix
Skip the averaged values and force top to show you the state of each logical processor individually so you can actually see which core is doing the heavy lifting.
top -1
⚙️ Why It Works
Passing the dash one flag forces the CPU summary to break out of the single aggregate line and display every core available on your hardware, which makes it dead easy to spot a single-threaded process thrashing one specific core.
🚀 Pro-Tip: Hit ‘1’ while top is running to toggle this view without restarting the process.
Linux Tips & Tricks | © ngelinux.com | 7/23/2026
