netstat Commands: A Comprehensive Guide
"Is anything listening on port 8080?" is a question you ask surprisingly often when running servers:
a port conflict, a service that failed to bind, or an unknown process holding a port. netstat
(Network Statistics) is the classic Linux tool for answering it. This guide covers the flags that matter,
shows real output so you know what you are looking at, and covers the errors you will meet — including the
one that happens the first time you try -p without root.
What netstat Can Tell You
- Which ports are open and listening, and which process owns them
- Active TCP connections (who is connected to your server right now)
- Protocol statistics (dropped packets, checksum errors)
- The kernel routing table and interface statistics
netstat is not always installed by default on modern systems; it is part of the
net-tools package. The newer alternative ss ships with iproute2 and is faster on
large outputs — see the comparison at the end.
# On Debian/Ubuntu systems
sudo apt-get install net-tools
# On RHEL/Rocky Linux/CentOS systems
sudo yum install net-tools
Essential netstat Commands
1. Find Which Process Is Listening on a Port
This is the command I run most: every TCP listener with its PID and program name.
netstat -nlpt
-n— numeric output, no slow DNS lookups or name resolution-l— only listening sockets-p— show the PID and program name (needs root to see other users' processes)-t— TCP only
Example output on a machine running Keycloak and PostgreSQL:
Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name
tcp 0 0 0.0.0.0:8080 0.0.0.0:* LISTEN 12879/java
tcp 0 0 127.0.0.1:5432 0.0.0.0:* LISTEN 1042/postgres
tcp 0 0 127.0.0.1:22 0.0.0.0:* LISTEN 890/sshd
0.0.0.0:8080 means the service listens on all interfaces — reachable from the network.
127.0.0.1:5432 means local connections only. If you expected a service to be reachable
remotely and it only shows on 127.0.0.1, that is your answer.
2. List All Listening TCP and UDP Ports
netstat -tuln
Adds -u for UDP. Useful for a quick "what is open on this box" audit.
3. Show Active (Established) Connections
netstat -tn
Hides listeners and shows live connections — handy when checking who is connected to your application.
4. Protocol Statistics
netstat -s
Per-protocol counters: packets sent/received, retransmissions, checksum errors, and dropped segments.
A growing Retransmitted packets count points to network loss or congestion.
5. Everything (Listeners + Connections + Unix Sockets)
netstat -a
6. Kernel Routing Table
netstat -r
Shows the default gateway and route rules. If your machine cannot reach the internet but networking is "up", check here that a default route exists.
7. Full Port Audit (TCP + UDP + PIDs)
netstat -tulpn
The standard "show me everything open on this box" command. The n keeps it fast on busy
servers.
8. Filter by IP Version
netstat -4 # IPv4 only
netstat -6 # IPv6 only
9. Map Connections to Program Names
netstat -tp
Same as -nlpt but without the listener filter — includes established connections.
Putting It Together: Diagnosing a Port Conflict
Suppose your new service refuses to start with "Address already in use" on port 8080. Here is the exact sequence I use:
# 1. Confirm what is on 8080
netstat -tulpn | grep 8080
# 2. Result: another java process owns it
tcp6 0 0 :::8080 :::* LISTEN 12879/java
# 3. Inspect that process to decide if it can be stopped
ps -fp 12879
You can then either stop the conflicting service or configure yours to use a different port.
Common Problems
Problem 1: "netstat: command not found"
Cause: net-tools is not installed on your distribution.
Solution: install it (commands above), or use the built-in ss:
ss -tulpn is the direct modern equivalent.
Problem 2: "You are not allowed to see the process info" or PIDs shown as "-"
Cause: -p needs root permission to see processes owned by other users.
Solution: run with sudo netstat -tulpn.
Problem 3: service listening but unreachable from outside
Cause: the process bound to 127.0.0.1 instead of 0.0.0.0, or a
firewall rule is blocking the port.
Solution: check the Local Address column in netstat -tulpn; then check
iptables/firewall rules and test the port with
netcat.
Problem 4: slow output / netstat seems stuck
Cause: without -n, netstat performs reverse DNS lookups on every address.
Solution: always include -n for production use.
netstat vs ss
| Feature | netstat | ss |
|---|---|---|
| Availability | net-tools (often needs install) | iproute2 (default on most distros) |
| Speed with many sockets | Slower | Faster |
| Key command | netstat -tulpn | ss -tulpn |
| Mature scripts | Still widely used in legacy scripts | Preferred for new work |
Best Practices
- Always use
-nto avoid slow DNS lookups. - Learn
ss -tulpnas the modern replacement; keep netstat knowledge for legacy servers. - When checking port conflicts, combine
netstatwithps -fp PIDto identify the offending process before killing anything. - On containers and busy hosts, prefer
ss— netstat can be slow with thousands of sockets.
Conclusion
netstat -tulpn answers most port questions in one line: what is listening, on which address,
and owned by which process. Read the Local Address column carefully (0.0.0.0 vs 127.0.0.1), remember that
-p wants root, and you will resolve the majority of binding and connection issues without
leaving the terminal.