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
A cluster network is a pod prefix, a service prefix, and a policy. The CIDR planner drafts those prefixes from sizes you type. They must not overlap the LAN you already use. Documentation ranges stay in the note.
A NetworkPolicy draft is a deny or an allow you can read. The default, if you have not set one, may be allow. Write which it is. The cloud desk owns the cluster product. This page owns the prefixes and one policy idea.
Do not apply a policy that locks out the only path you have to the API. A lab namespace is the place a policy is tried.
You do not probe a cluster you do not administer, and you do not use a policy generator to bypass someone else's control.
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.
- Name the cluster you operate.
- Write the pod and service prefixes.
- Check they do not overlap the LAN.
- Say whether the default policy allows.
- Draft one lab policy and do not apply it here.
- Keep the API path out of the experiment.
Mistakes that outlive the command
Kubernetes Networking Explained goes wrong in a small set of ways. The list is here so the ticket can forbid them before the window.
- Overlapping prefixes.
- A policy applied to the API by mistake.
- A cluster you do not run.
- A documentation prefix applied.
- A default-allow left unwritten.
Decisions to write down
One decision for Kubernetes Networking Explained: The cluster is yours. Prefixes do not overlap the LAN. The default is written. Nothing was applied. And the API path is safe.
Checks before you call it done
Confirm this before you call the change done: Both prefixes are in the ticket. Overlap was checked. The default policy is a word. No manifest was applied. And the cloud guide remains the cluster guide. If one of those fails, stop.
Prefixes that do not overlap the LAN.
Do not lock yourself out of the API.Before you leave the device
Kubernetes Networking 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.
Related pillar guides
More in this cluster
The pillar for this page, and the next published pages filed with it.
- Pillar: Data Center Networking Guide
- Explainer: Docker Networking Explained
- How-To: Connecting On-Prem to Cloud with VPN
- Explainer: Leaf-Spine Architecture Explained
Companion: Cloud: Kubernetes for Admins.