build(deps): bump github.com/jsimonetti/rtnetlink from 1.4.0 to 1.4.1

[dependabot]

Bumps github.com/jsimonetti/rtnetlink from 1.4.0 to 1.4.1.

Release notes

Sourced from github.com/jsimonetti/rtnetlink's releases.

v1.4.1

The behaviour of NeighAttributes has changed and length checking has been removed. This means you should check for correct lengths on the receiving end.

What's Changed

• Enable golangci-lint by @​SuperQ in jsimonetti/rtnetlink#202
• Use commit hash with actions by @​jsimonetti in jsimonetti/rtnetlink#203
• build(deps): bump actions/setup-go from 3.5.0 to 4.1.0 by @​dependabot in jsimonetti/rtnetlink#205
• build(deps): bump golang.org/x/sys from 0.14.1-0.20231108175955-e4099bfacb8c to 0.15.0 by @​dependabot in jsimonetti/rtnetlink#206
• build(deps): bump golang.org/x/sys from 0.15.0 to 0.16.0 by @​dependabot in jsimonetti/rtnetlink#210
• build(deps): bump actions/setup-go from 4.1.0 to 5.0.0 by @​dependabot in jsimonetti/rtnetlink#209
• Relieve length check for NeighAttributes by @​jsimonetti in jsimonetti/rtnetlink#208

Full Changelog: jsimonetti/rtnetlink@v1.4.0...v1.4.1

Commits

• 8f6dc23 Relieve length check for NeighAttributes (#208)
• 8740a9c build(deps): bump actions/setup-go from 4.1.0 to 5.0.0 (#209)
• 277e862 build(deps): bump golang.org/x/sys from 0.15.0 to 0.16.0 (#210)
• dbce757 build(deps): bump golang.org/x/sys (#206)
• 27a8b31 build(deps): bump actions/setup-go from 3.5.0 to 4.1.0 (#205)
• 13375a6 Use commit hash with actions (#203)
• 889c07b Enable golangci-lint (#202)
• 74450cb Revert #187
• See full diff in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

---

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot merge will merge this PR after your CI passes on it
• @dependabot squash and merge will squash and merge this PR after your CI passes on it
• @dependabot cancel merge will cancel a previously requested merge and block automerging
• @dependabot reopen will reopen this PR if it is closed
• @dependabot close will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore this major version will close this PR and stop Dependabot creating any more for this major version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this minor version will close this PR and stop Dependabot creating any more for this minor version (unless you reopen the PR or upgrade to it yourself)
• @dependabot ignore this dependency will close this PR and stop Dependabot creating any more for this dependency (unless you reopen the PR or upgrade to it yourself)

[SuperQ]

@jsimonetti do we need to make changes in the node_exporter for this? The change to the length check seems like it would be dangerous to ignore here. (IMO that was a bit of a breaking behavior change)

[jsimonetti]

For as far as I understood from @dswarbrick the arp collector already checks for the correct bounds and lengths.

The original check was actually too strict and blocked legitimate stuff. There are too many variances there to really make a decent check (tunnels, p2p, nbma, etc).

@dswarbrick did I understand you correctly?

[dswarbrick]

@jsimonetti Since the node_exporter code in question specifically replaces parsing of /proc/net/arp, it first and foremost filters out any rtnetlink messages which are not AF_INET, i.e., it ignores IPv6 neighbours and whatever else exotic protocols may be running on the host. Secondly, it skips entries which are NUD_NOARP, just as the kernel code which produces the human-readable /proc/net/arp contents does.

Beyond that, it doesn't care about the length of the link-layer addresses, since it merely counts the number of items per interface. It makes no attempt to parse the link-layer address byte slice back into a MAC address (or GUID, in the case of InfiniBand).

cf.

	ifIndexEntries := make(map[uint32]uint32)

	for _, n := range neighbors {
		// Neighbors will also contain IPv6 neighbors, but since this is purely an ARP collector,
		// restrict to AF_INET. Also skip entries which have state NUD_NOARP to conform to output
		// of /proc/net/arp.
		if n.Family == unix.AF_INET && n.State&unix.NUD_NOARP == 0 {
			ifIndexEntries[n.Index]++
		}
	}

• where n.Index is the interface index number. That's pretty much it.

[jsimonetti]

Thank you for your explanation! This means no changes are needed on the node_exporter end.