linuxexpert-practice-labs/labs/lpic-1/109-3-network-troubleshooting

LPIC-1 109.3: Basic network troubleshooting

Which process holds a port, a test connection with netcat through the firewall, and a fault found by working through the layers.

Before you start

The peer, at 10.99.0.2 (name peer), stands in for another VM. It sits on this VM's second network, 10.99.0.0/24, where this VM is 10.99.0.1 on interface le-host. Get a shell there with sudo ip netns exec lepeer bash. The lab has broken the link to the peer in one place, for task 3. Your own connection to the VM isn't touched.

Tasks

1. Find the process listening on port 22 with sudo ss -tlnp, and write its name to /root/lab-answers/1. Then install net-tools and compare with netstat -tlnp.
2. Install netcat if needed, listen on TCP port 9000 on this VM, and connect from the peer (nc 10.99.0.1 9000). Open the port in the firewall if one is running. Type in both directions. Then check the port from the peer with nc -zv, and leave a listener running on 9000 for the check (nc -lk 9000).
3. This VM can't reach the peer: ping 10.99.0.2 fails. Work through the troubleshooting order (link, address, route, then the far end) to find the fault, fix it, and write the name of the interface at fault to /root/lab-answers/3.

Hints

• ss -tlnp 'sport = :22'; on Ubuntu, systemd holds the port for ssh.socket. netstat is in net-tools.
• nc is in nmap-ncat on the RHEL family and netcat-openbsd on Ubuntu. firewall-cmd --add-port=9000/tcp (add --permanent to keep it), or ufw allow 9000/tcp. Run nc -lk 9000 & in the background, or under setsid.
• ip -br link, ip -br addr, ip route, ping.