Network Troubleshooting: A Practical Step-by-Step Method
Network connectivity issues are among the most common problems server administrators face. A systematic troubleshooting approach identifies the problem quickly rather than making random configuration changes.
Step 1: Verify Local Interface
Check that the network interface is up and has an IP address: ip addr show
Look for: - Interface state: should show UP - IP address: should be assigned and correct - No error flags on the interface
If the interface is down: sudo ip link set eth0 up If no IP is assigned: check DHCP or static configuration
Step 2: Test Local Connectivity
Ping the default gateway to verify local network connectivity: ping -c 3 gateway-ip
If this fails: - Verify the gateway IP is correct - Check ARP resolution: arp -n - Verify the interface subnet mask matches the gateway's subnet - Check for VLAN misconfiguration
Step 3: Test External Connectivity
Ping an external IP address (bypassing DNS): ping -c 3 8.8.8.8
If this works but DNS resolution fails, the problem is DNS-specific.
If this fails: - Check the routing table: ip route show - Verify the default route exists - Trace the packet path: traceroute 8.8.8.8 - Check firewall rules (iptables -L or nft list ruleset)
Step 4: Test DNS Resolution
Test name resolution: nslookup example.com or: dig example.com
If DNS fails: - Check /etc/resolv.conf for nameserver configuration - Try alternative DNS servers: dig @8.8.8.8 example.com - Verify DNS server reachability: ping nameserver-ip
Step 5: Test Application Connectivity
Test the specific application port: curl -v http://target-host:port or: ss -tnp | grep port-number
Check if the application is listening: ss -tlnp Check firewall rules for the specific port
Step 6: Check for Packet Loss and Latency
Extended ping test: ping -c 100 target-host
traceroute shows the path and per-hop latency: traceroute target-host
mtr combines ping and traceroute for continuous monitoring: mtr target-host
Step 7: Inspect Firewall Rules
Verify no firewall rules are blocking traffic: iptables -L -n -v (legacy) nft list ruleset (modern)
Check both input and output chains. Cloud platform security groups may also block traffic at the hypervisor level.
Common Issues
- Wrong default gateway: after server migration or network change
- DNS misconfiguration: /etc/resolv.conf overwritten by DHCP
- Firewall blocking: new rule added that blocks required traffic
- MTU mismatch: causes intermittent failures with large packets
- Routing loop: packets cycling between routers
Conclusion
Follow the OSI layers from bottom up: physical connectivity, IP configuration, routing, DNS, and finally application-layer issues. This systematic approach resolves most network problems efficiently.