Top 5 Terminal Monitoring Tools: htop, btop, glances, iotop & vmstat Compared

Why the Legacy top Command Is No Longer Enough

Every Linux distribution ships with the standard top utility. While functional, top lacks horizontal scrolling for long process arguments, provides no visual representation of per-core CPU load, cannot filter disk I/O per PID, and requires cryptic single-key shortcuts to kill runaway tasks.

When troubleshooting performance degradation on a production cloud VPS—such as high load averages, runaway PHP-FPM workers, rogue database queries, or disk I/O bottlenecks—having the right terminal telemetry HUD allows you to pinpoint the root cause within seconds rather than digging through raw text logs.

Tool 1: htop – The Gold-Standard Interactive Process Viewer

Written in C and relying on ncurses, htop is the universal replacement for top. It features full mouse support, color-coded per-CPU core meters, visual memory breakdown (resident vs cached), and interactive process tree viewing.

# Install htop on Ubuntu/Debian
sudo apt update && sudo apt install -y htop

# Launch htop
htop

Key Shortcuts: Press F5 for Tree View (parent-child relationships), F6 to sort by MEM% or CPU%, F9 to send SIGKILL/SIGTERM directly to a highlighted PID, and F4 to filter by name.

Tool 2: btop – Modern C++ Resource Monitor with Stunning Visuals

btop (and its predecessor btop++) is widely considered the most visually sophisticated terminal monitor in existence. Written in pure C++, it delivers rich ASCII/Unicode graphs for per-core frequencies, disk read/write throughput, download/upload network adapters, and running processes with zero configuration.

# Install btop
sudo apt install -y btop

# Launch btop
btop

Tool 3: iotop – Pinpointing Disk I/O Bottlenecks

High Linux load average is frequently caused not by CPU exhaustion, but by I/O wait (wa)—where processes freeze while waiting for slow disk read/write operations. iotop monitors Linux kernel I/O accounting to identify exactly which process is thrashing your disk.

# Install iotop
sudo apt install -y iotop

# Monitor only processes actively doing disk I/O (-o)
sudo iotop -o -d 1

This instantly exposes unindexed MySQL table scans or rogue backup scripts dumping hundreds of megabytes per second.

Tool 4: glances – Cross-Platform Monitor with Web UI & REST API

Developed in Python, glances monitors CPU, memory, network, disk, Docker containers, and system sensors. Its killer feature is running as a lightweight web server, allowing you to monitor your VPS from any web browser without maintaining an active SSH terminal.

# Install glances
sudo apt install -y glances

# Launch web server mode on port 61208
glances -w

Tool 5: vmstat – The Zero-Overhead Kernel Diagnostics Engine

When a server is under severe distress and heavier utilities cannot even launch, vmstat (Virtual Memory Statistics) provides instant kernel metrics on process run queues (r), blocked processes (b), swap in/out (si/so), and CPU context switches (cs).

# Display updates every 1 second
vmstat 1 10

Feature & Architecture Comparison Table

Tool	Primary Specialty	RAM Usage	Disk I/O Detail	Docker Container Support
htop	Process management & signaling	~4 MB	Basic	No
btop	Full visual hardware HUD	~14 MB	High (live graphs)	No
iotop	Disk I/O read/write per PID	~18 MB (Python)	Best in Class	No
glances	Remote Web UI & Prometheus export	~35 MB	Moderate	Yes (Native)
vmstat	Kernel memory & run queues	< 1 MB	Summary blocks	No

Frequently Asked Questions (FAQ)

Customizing htop Configuration: Creating Custom Column Profiles

Many sysadmins are unaware that htop settings can be persisted permanently by editing ~/.config/htop/htoprc. Customizing your default layout allows you to monitor critical operational metrics (such as CPU I/O wait, memory buffers, and resident set size) the moment you open the terminal:

# Recommended htop customizations in ~/.config/htop/htoprc
fields=0 48 17 18 38 39 40 2 46 47 49 1
sort_key=46
sort_direction=-1
tree_view=0
header_margin=1
detailed_cpu_time=1
show_cpu_frequency=1
show_cpu_temperature=1

Enabling detailed_cpu_time=1 differentiates between user processes, system kernel tasks, soft IRQs, and I/O wait with distinct color codes, allowing you to instantly determine whether high load is caused by database queries or hardware drive throttling.

Automating High Load Alert Scripts with Bash & Webhooks

While interactive terminal monitors are essential when logged in, production servers require automated alerting when CPU or memory thresholds are breached. You can deploy a lightweight bash watchdog script via crontab that triggers alerts to a Discord or Slack webhook:

#!/bin/bash
# Check 1-minute load average
LOAD=$(awk '{print $1}' /proc/loadavg | cut -d. -f1)
CORES=$(nproc)
THRESHOLD=$((CORES * 2))

if [ "$LOAD" -ge "$THRESHOLD" ]; then
  MESSAGE="⚠️ High Load Alert on $(hostname): Current load is ${LOAD} across ${CORES} CPU cores!"
  # Dispatch to webhook or email notification
  curl -s -X POST -H 'Content-type: application/json'     --data "{\"text\":\"$MESSAGE\"}" YOUR_WEBHOOK_URL
fi