gping — Ping with a Real-Time Terminal Graph

Introduction

gping takes the classic ping command and adds a live scrolling graph in the terminal. Each ping response plots as a point; latency spikes become obvious; packet loss shows as gaps. For anyone debugging network issues — ISP problems, Wi-Fi jitter, VPN overhead — the graph tells you in 10 seconds what a column of numbers takes minutes to convey.

With over 12,000 GitHub stars, gping is a must-have for network engineers, sysadmins, and anyone who's ever yelled at an unreliable Wi-Fi signal.

What gping Does

gping pings one or more targets at a fixed interval, renders a sliding-window graph on stdout using Unicode/ANSI drawing, and shows running stats (min/avg/max/jitter). The --cmd mode runs an arbitrary command periodically and graphs its runtime — useful for "how stable is this curl request?" scenarios.

Architecture Overview

[gping (Rust)]
      |
[Pinger]
   sends ICMP (needs raw socket perms) or TCP handshake
   parallel per target, 1s interval (configurable)
      |
[Buffer]
   ring buffer of (timestamp, latency) tuples
      |
[Graph Renderer]
   crossterm + tui-rs
   overlaid lines per target, colored
   auto-scaling Y axis
      |
[Stats panel]
   min / avg / max / jitter / loss per target

Self-Hosting & Configuration

# Useful flags
gping -n 10 host.example.com           # ping every 10s
gping --buffer 600 google.com          # widen X axis to 10 min
gping -4 google.com                    # IPv4 only
gping -6 google.com                    # IPv6 only

# TCP-mode (uses TCP handshakes instead of ICMP)
# Helpful when ICMP is filtered by firewalls
gping --tcp github.com:443

# Compare network + endpoint latency
gping --cmd "curl -s https://api.example.com/health -o /dev/null" \
      --cmd "curl -s https://api-backup.example.com/health -o /dev/null"

# Show historical stats when quitting
# Press q to quit; summary prints

Key Features

• Real-time terminal graph — Unicode-drawn sliding window
• Multi-target overlay — compare several endpoints simultaneously
• TCP mode — works where ICMP is filtered
• Command mode — graph any command's runtime, not just ping
• Auto-scaling axis — zooms Y axis to fit latest values
• Jitter + loss stats — real-time summary alongside the graph
• Cross-platform — Rust binary on macOS, Linux, Windows
• Color-coded per target — distinguish overlays at a glance

Comparison with Similar Tools

FAQ

Q: Why not just use mtr? A: mtr shows hop-by-hop latency (useful for "where is the problem?"). gping shows end-to-end latency over time (useful for "when is it slow?"). Different questions; use both.

Q: Does gping need root? A: ICMP mode usually yes (raw sockets). On Linux, set capability setcap cap_net_raw+ep $(which gping) once. TCP mode (--tcp) works without privileges.

Q: Can I use it in tmux or screenshots? A: Yes — the Unicode output looks fine in any modern terminal. It's a popular share-a-screenshot-on-Twitter tool when debugging ISP issues.

Q: Does it log to a file? A: Not directly. For historical data, pair with a monitoring tool (Smokeping, Prometheus blackbox exporter) — gping is for interactive debugging.

Sources

• GitHub: https://github.com/orf/gping
• License: MIT