ip route Command in Linux: Syntax, Options & Routing Examples

Deepak Prasad
Tested on Ubuntu 25.04 (Plucky Puffin)
Package iproute2 6.14.0
iproute2 6.14.0
Applies to Ubuntu, Debian, RHEL, Fedora
Privilege sudo or root
Man page ip route(8)
Scope ip route displays and changes the kernel routing table — default gateways, static subnets, metrics, and special drop routes. It is part of iproute2 and replaces the legacy route command on modern Linux.
Related guides ss
Linux commands

ip route — quick reference

View routes

Inspect what the kernel will use before you change anything.

When to use Command
Show the main routing table (same as show) ip route
Show routes explicitly ip route show
Show only the main table ip route show table main
Filter routes on one interface ip route show dev enp0s3
Show default gateway routes ip route show default
List routes in every table ip route show table all
Ask which path the kernel picks for a destination ip route get 1.1.1.1

Add and change routes

Static routes added this way are temporary until reboot unless you persist them in Netplan, NetworkManager, or similar.

When to use Command
Add a subnet via a gateway sudo ip route add 192.168.99.0/24 via 10.0.2.2 dev enp0s3
Add a directly connected subnet on an interface sudo ip route add 172.16.50.0/24 dev enp0s8
Set route priority with a metric (lower wins) sudo ip route add 192.168.99.0/24 via 10.0.2.2 dev enp0s3 metric 200
Update or create a route in one step sudo ip route replace 192.168.99.0/24 via 10.0.2.2 dev enp0s3 metric 150
Pin a preferred source address on a route sudo ip route add 192.168.99.0/24 via 10.0.2.2 src 10.0.2.15

Delete and flush

When to use Command
Remove a route to a prefix sudo ip route del 192.168.99.0/24
Remove a default route sudo ip route del default
Drop the routing cache (after topology changes) sudo ip route flush cache

Special route types

These routes control traffic without normal forwarding — useful for blacklisting subnets.

When to use Command
Silently discard packets to a network sudo ip route add blackhole 192.168.200.0/24
Drop and send ICMP unreachable sudo ip route add unreachable 192.168.201.0/24
Block with ICMP administratively prohibited sudo ip route add prohibit 192.168.202.0/24

ip route — command syntax

Synopsis from ip route help on Ubuntu 25.04 (iproute2 6.14.0):

text
ip route { list | show | flush } SELECTOR
ip route get [ ROUTE_GET_FLAGS ] [ to ] ADDRESS
       [ from ADDRESS iif STRING ] [ oif STRING ] ...
ip route { add | del | change | append | replace } ROUTE
SELECTOR := [ root PREFIX ] [ match PREFIX ] [ table TABLE_ID ] ...
ROUTE := NODE_SPEC [ INFO_SPEC ]
NODE_SPEC := [ TYPE ] PREFIX [ via ADDRESS ] [ dev STRING ] [ metric METRIC ] ...
TYPE := { unicast | blackhole | unreachable | prohibit | ... }
TABLE_ID := [ local | main | default | all | NUMBER ]

ip route updates the kernel routing table in memory. Changes from add, del, and replace are not saved across reboot unless you configure your distro's network stack. Most write operations need sudo. See the broader ip command for addresses, links, and rules.

ip route — command examples

Essential Show the routing table

Inspect the routing table before changing anything — you need to know current gateways and interfaces.

Run the command:

bash
ip route show

Sample output:

output
default via 10.0.2.2 dev enp0s3 proto dhcp src 10.0.2.15 metric 100
default via 192.168.0.1 dev enp0s8 proto dhcp src 192.168.0.4 metric 101
10.0.2.0/24 dev enp0s3 proto kernel scope link src 10.0.2.15 metric 100
192.168.0.0/24 dev enp0s8 proto kernel scope link src 192.168.0.4 metric 101

Each line is one route: default is the catch-all, via is the next-hop gateway, and dev is the outgoing interface. When two default routes exist, the lower metric wins (here enp0s3 at 100 beats enp0s8 at 101).

Essential See which route the kernel picks

When traffic fails or takes an unexpected path, ip route get simulates the forwarding decision for one destination.

Run the command:

bash
ip route get 1.1.1.1

Sample output:

output
1.1.1.1 via 10.0.2.2 dev enp0s3 src 10.0.2.15 uid 0
    cache

The kernel would send packets to 1.1.1.1 through gateway 10.0.2.2 on enp0s3, using source address 10.0.2.15. Compare this with ip route show default when you suspect the wrong gateway is active.

Common Filter by interface and default route

On multi-homed hosts, narrow the table to one NIC or check only default routes.

Run the commands:

bash
ip route show dev enp0s3
ip route show default

Sample output:

output
default via 10.0.2.2 proto dhcp src 10.0.2.15 metric 100
10.0.2.0/24 proto kernel scope link src 10.0.2.15 metric 100
default via 10.0.2.2 dev enp0s3 proto dhcp src 10.0.2.15 metric 100
default via 192.168.0.1 dev enp0s8 proto dhcp src 192.168.0.4 metric 101

The first block lists routes tied to enp0s3; the second shows every default gateway. Use this when one interface should carry internet traffic and the other is internal-only.

Common Add a temporary static route

Reach an internal subnet through a router on your LAN. Routes added with ip route add disappear after reboot unless you persist them in Netplan or NetworkManager.

Run the command:

bash
sudo ip route add 192.168.99.0/24 via 10.0.2.2 dev enp0s3 metric 200

Verify it landed:

bash
ip route show 192.168.99.0/24

Sample output:

output
192.168.99.0/24 via 10.0.2.2 dev enp0s3 metric 200

Remove the test route when you are done:

bash
sudo ip route del 192.168.99.0/24

Always delete lab routes so the host returns to its original table.

Common Replace a route without a separate delete

When a gateway or metric changes, replace updates the matching prefix or creates it if missing.

Run the commands:

bash
sudo ip route add 192.168.99.0/24 via 10.0.2.2 dev enp0s3 metric 200
sudo ip route replace 192.168.99.0/24 via 10.0.2.2 dev enp0s3 metric 150
ip route show 192.168.99.0/24
sudo ip route del 192.168.99.0/24

Sample output after replace:

text
192.168.99.0/24 via 10.0.2.2 dev enp0s3 metric 150

replace is safer than del followed by add when scripts update routes in place.

Common Route a subnet out an interface without via

When the destination network is directly attached, specify dev instead of a gateway.

Run the commands:

bash
sudo ip route add 172.16.50.0/24 dev enp0s8
ip route show 172.16.50.0/24
sudo ip route del 172.16.50.0/24

Sample output:

output
172.16.50.0/24 dev enp0s8 scope link

The kernel treats the subnet as on-link through enp0s8 — no via gateway is needed.

Advanced Blackhole, unreachable, and prohibit routes

Special route types drop or reject traffic without a firewall rule. blackhole is silent; unreachable and prohibit return ICMP errors.

Run the commands:

bash
sudo ip route add blackhole 192.168.200.0/24
sudo ip route add unreachable 192.168.201.0/24
sudo ip route add prohibit 192.168.202.0/24
ip route show | grep 192.168.20
sudo ip route del blackhole 192.168.200.0/24
sudo ip route del unreachable 192.168.201.0/24
sudo ip route del prohibit 192.168.202.0/24

Sample output:

output
blackhole 192.168.200.0/24
unreachable 192.168.201.0/24
prohibit 192.168.202.0/24

Use these for coarse traffic blocks at the routing layer; remove them after testing.

Advanced Flush the routing cache

After link or gateway changes, stale cache entries can mislead ip route get. Flushing the cache forces fresh lookups.

Run the command:

bash
sudo ip route flush cache

There is no output on success. Re-run ip route get for the destination you care about to confirm the kernel picked the new path.

ip route — when to use / when not

Use ip route when Use something else when
You need to view or change the kernel routing table on a running Linux host You only need to ping a host — use ping first
You are on a distro that ships iproute2 (default on Ubuntu 25.04) You are on a legacy script that still calls route from net-tools — migrate to ip route
You want temporary static routes for testing or automation Routes must survive reboot — configure Netplan, NetworkManager, or systemd-networkd instead
You want to simulate which path a packet takes You need hop-by-hop path discovery — use traceroute

ip route vs route

ip route (iproute2) route (net-tools)
Package iproute2 net-tools (deprecated)
Maintenance Actively maintained Legacy
Policy routing Yes (table, works with ip rule) No
Multipath / nexthop Yes Limited
Output Structured, script-friendly Older column layout
Modern default Yes on Ubuntu, Debian, RHEL 8+ Only if net-tools is installed

Example equivalents:

bash
ip route show
route -n
bash
sudo ip route add default via 10.0.2.2
sudo route add default gw 10.0.2.2

Prefer ip route for new scripts and documentation.

ip route — interview corner

Practice these before exams or standups. Each card explains the idea in plain language, then ends with a short answer you can say aloud.

What does ip route do in Linux?

ip route is the iproute2 subcommand that reads and writes the kernel routing table. It answers two everyday questions: where should packets for this destination go, and how do I add or remove a static path?

View the table:

bash
ip route show

Simulate one forwarding decision:

bash
ip route get 1.1.1.1

It replaces the older route command from net-tools and works alongside ip addr, ip link, and ip rule in the ip command suite.

A strong answer is:

"ip route manages the kernel routing table — show, add, delete, and replace routes. ip route get tells me which gateway and interface the kernel would use for a destination. It is part of iproute2 and replaces legacy route on modern Linux."

What is the difference between ip route show and ip route get?

ip route show lists configured routes — everything the admin (or DHCP) installed in the table.

ip route get simulates forwarding for one destination address. It reports the chosen gateway, interface, and source IP the kernel would use right now, including cache state.

bash
ip route get 1.1.1.1

Sample output:

output
1.1.1.1 via 10.0.2.2 dev enp0s3 src 10.0.2.15 uid 0
    cache

Use show for inventory; use get when a host is reachable but traffic takes the wrong NIC or gateway.

A strong answer is:

"show lists routes in the table; get simulates where one packet would go — gateway, dev, and src. I use get when the table looks fine but actual traffic picks the wrong path."

What does metric mean on an ip route?

Metric is route priority when multiple routes match the same destination. Lower metric wins.

On a dual-homed host you might see:

text
default via 10.0.2.2 dev enp0s3 metric 100
default via 192.168.0.1 dev enp0s8 metric 101

Traffic uses 10.0.2.2 on enp0s3 because 100 < 101. Raise the backup link's metric to keep it idle until the primary fails (depending on your routing setup).

A strong answer is:

"Metric is route priority — lower number wins when several routes match. I set a low metric on the primary default gateway and a higher one on the backup NIC."

Are ip route changes permanent?

No. Routes added with ip route add, replace, or del live in the running kernel only. They disappear after reboot or sometimes after a network restart.

For persistence on Ubuntu, put routes in Netplan (/etc/netplan/*.yaml) or NetworkManager profiles with the nmcli command (connection modify … +ipv4.routes). Use ip route for temporary routes and boot-time scripts managed by your network stack.

A strong answer is:

"ip route changes are runtime-only. For persistence I configure Netplan or NetworkManager; ip route is for inspection and temporary static routes."

When would you use a blackhole route?

A blackhole route silently drops packets to a prefix — no ICMP reply. It is a routing-layer block, not a firewall rule.

bash
sudo ip route add blackhole 192.168.200.0/24

Use cases include sinking unwanted traffic to unused RFC1918 space or absorbing probes to decommissioned subnets. unreachable and prohibit differ because they return ICMP errors to the sender.

A strong answer is:

"Blackhole routes discard traffic to a prefix without notifying the sender — useful for sinkholing unwanted destinations. Unreachable and prohibit also block but send ICMP errors back."

Troubleshooting

Symptom Likely cause Fix
RTNETLINK answers: File exists on add Duplicate route for the same prefix ip route show then sudo ip route del PREFIX or use replace
RTNETLINK answers: No such process on del Selector does not match exactly (metric, via, dev) Match the line from ip route show in the del command
No default route in ip route show default DHCP failure or manual config removed gateway Add with sudo ip route add default via GATEWAY dev IFACE; fix Netplan/NM for persistence
Wrong interface for internet traffic Lower metric on the wrong default route Adjust metrics or remove the unwanted default
ip route get shows unexpected src Multiple addresses on one host Add src ADDRESS on the route or use policy routing with ip rule
Routes vanish after reboot Only runtime ip route changes were made Persist in Netplan, NetworkManager, or cloud-init
Rohan Timalsina

is a technical writer and Linux enthusiast who writes practical guides on Linux commands and system administration. He focuses on simplifying complex topics through clear explanations.

Expertise
  • Linux
  • HTML5
  • JavaScript
  • Web Design
  • Front-end Web Development

Related Articles

cheatsheet

traceroute Command in Linux: Syntax, Options & Route Tracing Examples

traceroute prints each router hop between your Linux host and a destination by sending probes with increasing …

11 min readcheatsheet

ip Command in Linux: Addresses, Links, Routes, and Troubleshooting

ip from iproute2 configures addresses, links, and routes on Linux. Subcommands ip addr, ip link, and ip route …

9 min readlinux

Troubleshoot Default Routes in Linux

Troubleshoot Linux default route problems by reading ip route show default and ip route get, understanding …

9 min read