A Minimal Triage Routine for “It’s Slow”

2026-02-10 · Linux · Ops

When something feels slow, the goal is to find the bottleneck, not to guess. Here is a small routine that catches most issues quickly.

1) CPU and load

top -o cpu
uptime

2) Memory pressure

vm_stat
sysctl vm.swapusage

3) Disk and IO

df -h
iostat -w 1

4) DNS

dig +short example.com
scutil --dns | head

5) Routing / reachability

ping -c 3 1.1.1.1
traceroute 1.1.1.1

The win is consistency. If you always check the same items in the same order, you stop chasing noise.