Back to Articles
Cloud & Network7 min readHow Internet Requests Actually Work · Part 8 of 8

Practical Network Troubleshooting: Ping, DNS, Ports, Cloud, and Containers

Use ping, dig, curl, traceroute, route, and socket tools correctly; distinguish latency from bandwidth; and apply the same networking fundamentals to localhost, cloud networks, Docker, and Kubernetes.

TroubleshootingPingTracerouteCloud NetworkingContainers

Troubleshooting becomes easier when every tool is tied to a layer. No single command proves that an application works, and a failed ping does not prove a server is offline.

Diagram 1

Networking Troubleshooting Basics

Use the diagnostic tool that tests the layer connected to the symptom instead of treating every failure as the same network problem.

Test in Layer Order

Resolve name -> Inspect route -> Reach port -> Negotiate TLS -> Send HTTP -> Inspect application

dig / nslookup
Inspect DNS answers.
ping
Uses ICMP echo to test basic reachability when replies are allowed.
traceroute / tracert
Uses hop-limit behavior to reveal responding hops.
curl
Tests HTTP, redirects, TLS, headers, and response details.
ip addr / ipconfig
Shows local interfaces and addresses.
ip route / route print
Shows host routes and gateways.
ss / netstat
Shows listening sockets and connections.

ICMP may be blocked while HTTPS works. Ping may succeed while the web process is stopped. Test the layer connected to the symptom.

Latency, Bandwidth, and Loss

Bandwidth describes how much data a path can carry over time; latency describes how long data takes to travel. A high-bandwidth path can still feel slow for chatty applications when every round trip has substantial delay.

Packet loss can come from congestion, wireless interference, damaged links, overloaded equipment, or policy. TCP retransmits missing data, protecting correctness but increasing delay. Real-time UDP applications may conceal or tolerate loss.

Localhost and Listening Addresses

localhost normally resolves to loopback such as 127.0.0.1 or ::1. A service bound to 127.0.0.1:3000 is reachable only through that machine's loopback interface. Binding to 0.0.0.0:3000 listens on all IPv4 interfaces inside the process's network namespace.

Listening on 0.0.0.0 does not make a service public. Routes, firewalls, security groups, NAT, container port publishing, and load balancers still decide who can connect.

Common container surprise

Bind 127.0.0.1 in container -> only that container reaches itBind 0.0.0.0 -> container interfaces can accept trafficHost or Kubernetes networking must still publish and route the port

Cloud and Containers Use the Same Foundations

A cloud design may expose a public load balancer while keeping applications and databases private. Policies allow only required flows. Docker and Kubernetes add namespaces, virtual interfaces, service discovery, overlays, Services, and ingress, but they do not replace IP, ports, DNS, routes, TCP, UDP, NAT, or firewalls.

Internet -> Public load balancer -> Private application -> Private database

Diagram 2

Real Networking Examples

Websites, APIs, streams, and IoT traffic share the same addressing and routing foundation while making different transport trade-offs.

A Repeatable Diagnostic Habit

Write down the expected hostname, resolved address, destination port, transport, client network, server listener, and every policy boundary. Test them in order. This turns the network is down into a specific observation another engineer can reproduce.

Final distinction

A request is an application action, a connection is transport state, and a packet carries part of the communication. Good troubleshooting identifies which one failed.

Series

How Internet Requests Actually Work

An eight-part beginner series tracing internet communication from an HTTP request through DNS, IP, ports, transport protocols, packets, network devices, server infrastructure, security, and troubleshooting.

Part 8 of 8
About the Author

Nikhlesh Yadav is a Technical Lead and Solution Architect with 12+ years of experience across cloud-native systems, distributed platforms, AI integrations, Web3, and cyber security.

Read full profile