When experiencing intermittent latency or disconnections, start with these basic diagnostics:

# Continuous ping test to multiple targets
ping -t google.com &> google_ping.log &
ping -t 8.8.8.8 &> dns_ping.log &
ping -t your_router_ip &> local_ping.log

Mapping the network path reveals where packets are failing:

# Windows
tracert google.com

# Linux/Mac
traceroute -n google.com

# MTR (combines ping+traceroute)
mtr --report-wide --report-cycles=10 google.com

Eliminate DNS as a potential culprit:

# Compare ISP DNS with public alternatives
nslookup example.com  # Uses default DNS
nslookup example.com 8.8.8.8  # Google DNS
nslookup example.com 1.1.1.1  # Cloudflare DNS

Measure actual throughput and packet loss:

# Speedtest CLI (install via pip)
speedtest-cli --secure --single

# iPerf3 for local network testing
# On server: iperf3 -s
# On client: iperf3 -c server_ip -t 60 -P 8

Check for local interface errors that might cause problems:

# Linux
ifconfig | grep -i errors
ip -s link show eth0

# Windows PowerShell
Get-NetAdapterStatistics | Select-Object Name,ReceivedErrors,SentErrors

Provide these diagnostic results when contacting support:

# Sample evidence package for ISP
ping -n 100 8.8.8.8 > ping_results.txt
tracert 8.8.8.8 > tracert_results.txt
netsh interface ip show config > network_config.txt

As developers, we often face connectivity issues that could stem from either our local setup or the ISP. Before reaching for the phone, let's approach this systematically with code and command-line tools.

# Continuous ping test to multiple targets
ping -t 8.8.8.8 | tee google_ping.log &
ping -t 192.168.1.1 | tee router_ping.log &

Run these simultaneously in separate terminals. Comparing results helps identify where packet loss occurs.

# Windows:
tracert 8.8.8.8

# Linux/macOS:
traceroute -n 8.8.8.8

# Python alternative:
import subprocess
def run_traceroute(target):
    try:
        result = subprocess.run(['traceroute', '-n', target], 
                              capture_output=True, text=True)
        return result.stdout
    except:
        return "Traceroute failed"

Create a simple Python script to measure actual throughput:

import speedtest
import time

def test_speed():
    st = speedtest.Speedtest()
    st.get_best_server()
    
    print("Testing download speed...")
    download = st.download() / 10**6  # Convert to Mbps
    print(f"Download: {download:.2f} Mbps")
    
    print("Testing upload speed...")
    upload = st.upload() / 10**6
    print(f"Upload: {upload:.2f} Mbps")
    
    return download, upload

while True:
    dl, ul = test_speed()
    time.sleep(300)  # Run every 5 minutes

DNS issues often masquerade as connectivity problems:

# Compare different DNS providers
dig @8.8.8.8 example.com
dig @1.1.1.1 example.com
dig example.com  # Your default DNS

# Python DNS lookup:
import socket
def check_dns(hostname, dns_server='8.8.8.8'):
    resolver = dns.resolver.Resolver()
    resolver.nameservers = [dns_server]
    try:
        answers = resolver.resolve(hostname)
        return [str(rdata) for rdata in answers]
    except Exception as e:
        return f"DNS query failed: {e}"

When dealing with intermittent latency, packet-level analysis helps:

# tcpdump example (Linux/macOS)
tcpdump -i eth0 -w packet_capture.pcap host 8.8.8.8

# Wireshark filter syntax for latency analysis:
# (tcp.analysis.ack_rtt > 0.1) && !(http or dns)

Automate monitoring with this Python snippet:

import requests
import ping3
import matplotlib.pyplot as plt
from datetime import datetime

def network_health_check():
    results = {
        'timestamp': datetime.now(),
        'google_ping': ping3.ping('8.8.8.8', unit='ms'),
        'cloudflare_ping': ping3.ping('1.1.1.1', unit='ms'),
        'http_status': requests.get('https://example.com').status_code
    }
    return results

# Store results in time-series database or CSV for trend analysis

By implementing these techniques, you'll have concrete data to present to your ISP when needed, turning subjective complaints into actionable technical reports.