From cb9b4fb30b49fd0b820e8f20c3ce8dad97e65370 Mon Sep 17 00:00:00 2001 From: Mark Petrovic Date: Thu, 19 Oct 2017 11:15:33 -0700 Subject: [PATCH] Update scratch.md (#5081) This document is very good, but it does not mention anywhere how a node would obtain its own Pod CIDR when setting up static pod routing on something like bare metal. I also think the kube-proxy section on this page should include *--cluster-cidr* in the "arguments to consider" area. I say should because I'm not confident that I'm right. If I omitted --cluster-cidr from the kube-proxy and controller manager, and found that my pod routing didn't work, adding the switch would be one of the first things I'd try. --- docs/getting-started-guides/scratch.md | 1 + 1 file changed, 1 insertion(+) diff --git a/docs/getting-started-guides/scratch.md b/docs/getting-started-guides/scratch.md index d03e813199..97b8ad5f59 100644 --- a/docs/getting-started-guides/scratch.md +++ b/docs/getting-started-guides/scratch.md @@ -407,6 +407,7 @@ Arguments to consider: - `--cluster-domain=` to the dns domain prefix to use for cluster DNS addresses. - `--docker-root=` - `--root-dir=` + - `--pod-cidr=` The CIDR to use for pod IP addresses, only used in standalone mode. In cluster mode, this is obtained from the master. - `--configure-cbr0=` (described below) - `--register-node` (described in [Node](/docs/admin/node/) documentation.)