atop — Full-system resource monitor

Linux / Service and process management / atop.html

Records CPU, memory, disk and network usage together and can replay past intervals.

Basic Information

What is this command?

atop shows per-process usage together with system-wide counters in one screen.

Its interval log lets you replay the exact state of the machine around an incident.

What problems does it solve?

• Find which process caused a short load spike.
• Decide whether memory, disk or network is the bottleneck.
• Replay the machine state at the moment of an incident.
• Record at intervals overnight and review the trend in the morning.

Key features

Live view

The default run refreshes system counters and top processes at a fixed interval.

• When to use: When you need a quick current picture
• What you get: CPU, memory, disk and network summary
• Limitations: Short spikes may fall between intervals

Interval logging and replay

Write samples to a file at a fixed interval and replay them with -r.

• When to use: When investigating after an incident
• What you get: Per-interval counters and process list
• Limitations: Log files grow, so a retention policy is needed

Switching process views

Interactive keys change sorting and open detailed per-process views.

• When to use: When drilling into one process
• What you get: Sorted process list with detail
• Limitations: Requires learning the key bindings

Installation and usage

Check whether it is installed

atop -V

apt install atop # Debian/Ubuntu dnf install atop # Fedora/RHEL

Real-world examples

Live monitoring

Refresh every 5 seconds.

# Refresh every 5 seconds
atop 5

Interval logging

Log one day of samples every 60 seconds.

# Sample every 60 seconds into a log file
atop -w /var/log/atop.daily 60

Replaying a log

Replay a recorded file from the start.

# Replay the recorded samples
atop -r /var/log/atop.daily