Practical tools for system, network, security, and operations administrators.

NET-C0112 · Routing

Policy-Based Routing Explained

Draft policy-based routing for one flow on a router you administer, and do not use it to evade a network you do not run.

Updated 25 October 2026 · Explainer

This is an original planning guide. It is not a vendor manual, an official standard, or legal advice. Example addresses use documentation ranges. Review every command before you paste it.

What this hour is for

Policy-based routing sends matching traffic to a next hop you choose, instead of the ordinary table. The builder drafts a stanza. It does not configure a router. The match has to be a source or a service you can say. 0.0.0.0/0 as the match is a line you delete.

The next hop is a router you administer or the provider hop you were given. A policy whose purpose is to get around a control on a network you do not administer is not this page. One policy.

Do not apply it during the reading.

The change you can finish

Apply nothing during the writing of the note. When the lab window opens, it is one change on the device named in the ticket. Read the draft, then the device, then decide.

Rollback is the file or the revision you saved first. You do not invent a rollback after the only copy is gone.

  1. Name the router.
  2. Write the match.
  3. Write the next hop.
  4. Delete a match of every address.
  5. Do not apply the policy.
  6. Stop.

Mistakes that outlive the command

Policy-Based Routing Explained goes wrong in a small set of ways. The list is here so the ticket can forbid them before the window.

Decisions to write down

One decision for Policy-Based Routing Explained: The router is yours. The match is specific. The next hop is known. Nothing was applied. And no evasion is the goal.

Checks before you call it done

Confirm this before you call the change done: The match is written. The next hop is written. A wide match is gone. No policy was applied. And the router was left alone. If one of those fails, stop.

Before you leave the device

Policy-Based Routing Explained ends when the hostname on the device matches the hostname in the ticket. If it does not, you are on the wrong box. Stop. A command that succeeded on the wrong box is not progress.

The draft stays a draft until a person has read the names in it. A generator does not log into the device and does not know which interface is the uplink. You delete a line that names a device, a zone, or a prefix you did not write down.

0.0.0.0/0 is not kept because a sample included it. A source you can say is the source you keep. 192.0.2.0/24, 198.51.100.0/24, 203.0.113.0/24, and 2001:db8::/32 are documentation. They belong in the note. They are not prefixes you type onto a production device because they appeared in a draft.

prompt-for-this is a reminder that a secret is missing. It is not a password, a key, a community string, or a pre-shared key you save in the ticket or in git. The real secret stays in the store.

One change is the window. A VLAN and a routing protocol and a VPN do not share it. If two things move and the result is wrong, you will not know which one did it. The next feature waits for its own page.

If the same step fails twice, stop for the day. The failure is the result. Adding a forum command beside a half-applied config produces a device nobody can rebuild. Restore the copy you took first.

A second person reads the device name and the change aloud. If they cannot find them in the ticket, the note is not done. Silence is not agreement. Then leave the device alone.

This page is a planning guide. It is not a vendor manual, not an official standard, and not a permission to touch a network you do not administer. The pillar guide owns the wider design. This page owns one lab proof.

Tools for this guide

These tools run in the browser on this site. They do not log into your machines.

More in this cluster

The pillar for this page, and the next published pages filed with it.