Linux Monitoring TTY

tload — terminal load-average graph

tload draws an ASCII graph of the system load average in your terminal. It’s lightweight, old-school, and great for quick “is this box busy?” checks over SSH.

What tload is good for

Quick visual check of system load in a plain terminal
Watching load spikes while running a build or backup
Low-overhead monitoring on older or minimal systems
Teaching/understanding load average at a glance

Tip: “Load average” is not CPU percent. It roughly represents how many tasks are competing for CPU (and sometimes uninterruptible I/O) over time.

Install / availability

On many distros, tload is provided by the procps / procps-ng utilities bundle. If tload isn’t found, install that package with your distro’s package manager.

# Debian/Ubuntu (common)
sudo apt update
sudo apt install procps

# RHEL/CentOS/Fedora (common)
sudo dnf install procps-ng

# Arch (common)
sudo pacman -S procps-ng

10 Practical Examples

1) Run tload with defaults

tload

Shows an updating load-average graph using the default refresh interval.

2) Refresh faster (every 1 second)

tload -d 1

-d sets the delay between updates in seconds.

3) Refresh slower (every 5 seconds)

tload -d 5

Useful when you just want a gentle trend line without constant redraw.

4) Fit the graph to a specific width

tload -w 80

Force the graph width (handy in split panes or narrow terminals).

5) Fit the graph to a specific width and refresh speed

tload -w 120 -d 2

Good for wide terminals: readable graph with a calm refresh rate.

6) Run tload in a tiny corner pane (tmux)

# In tmux, split a small pane, then run:
tload -w 40 -d 2

Great “always-on” load watch while you work in other panes.

7) Watch load while doing something heavy

# Terminal 1
tload -d 1

# Terminal 2 (example heavy work)
make -j$(nproc)

Quick sanity check: did your build or job push the system into high load?

8) Exit cleanly

# Press:
Ctrl+C

Use Ctrl+C to stop tload.

9) Compare with uptime

# Terminal 1
tload -d 1

# Terminal 2
uptime

uptime prints the 1, 5, and 15 minute load averages numerically.

10) Follow up when load is high: identify the cause

# If tload looks high, follow up with:
top
# or (often nicer):
htop

tload answers “busy or not?” — then top/htop answers “what’s doing it?”

Notes & Gotchas

Not CPU%: load average is about runnable (and sometimes uninterruptible) tasks, not direct CPU usage.
Multi-core matters: a load of 4 on a 4-core box may be “fine”; on a 1-core box, it’s “busy.”
SSH friendly: works great in plain terminals where GUI monitoring isn’t available.
Minimalist by design: no per-process info — pair it with top when needed.