uptime

Overview

uptime shows how long the system has been running, how many users are logged in, and the load average (1, 5, and 15 minutes). It is a quick health snapshot before diving into top/htop/vmstat.

Load average is the average number of runnable plus uninterruptible (disk sleep) tasks — not a pure “CPU %” meter.

Syntax

uptime [options]

Common Options

Option	Description
-p, --pretty	Pretty duration only
-s, --since	System up since timestamp
-h, --help	Help
-V	Version

Examples with Explanations

Default

uptime
# 10:15:02 up 21 days,  4:03,  3 users,  load average: 0.08, 0.12, 0.09

Pretty / since

uptime -p
uptime -s

Interpret load vs CPUs

uptime
nproc
lscpu | grep '^CPU(s):'
# rough rule of thumb: load near nproc ≈ busy; much higher may mean saturation

Watch over time

watch -n1 uptime

Scripts

load=$(awk '{print $1}' /proc/loadavg)
echo "1-minute load: $load"
# /proc/loadavg is the source of truth
cat /proc/loadavg

Correlate with other tools

uptime
who
w
top -bn1 | head

Containers

uptime
# may reflect host uptime depending on visibility of /proc

Understanding load average

Observation	Possible meaning
Low load, high latency	Network/app issues, not CPU
Load ≫ nproc, CPU idle% high	I/O wait / uninterruptible tasks
Load ≈ runnable tasks	CPU contention
Sudden load spikes	Cron storms, backups, build jobs

Use vmstat 1, iostat -xz 1, ps, pidstat for root cause.

Notes / Pitfalls

• Load is not “percent CPU”; a 64-core host can show load 20 and still be fine.
• Suspend-to-RAM / cloud live migration can make “up” time less intuitive.
• User count is from utmp; containers/ssh multiplexing may look odd.
• Don’t alert only on load without CPU, memory, and I/O context.

2026-relevant notes

• For long-term trends use Prometheus node exporter / sar rather than one-off uptime.
• Latency-sensitive systems care more about p99 and steal time (VMs) than raw load.
• systemd-analyze helps boot time; uptime -s shows when the current boot began.

Related Commands

• w / who — who is logged in
• top / htop / btop — live process view
• vmstat / iostat — run queue vs block I/O
• nproc / lscpu — CPU count
• cat /proc/loadavg — raw values
• sar — historical load

Additional Resources

• man uptime, man 5 proc (/proc/loadavg)