Why router IPs are considered unreachable?

[ccll]

Hi there, I've currently hit by a bug in another software (Gitpod) which uses 'is-reachable' to test the reachability of hosts, digging into the source codes leads me here, I'm not sure if it's relevant so I opened up an issue to discuss it a bit, feel free to close it if inappropriate.

The situation is one of my services is hosted in a home lab setup which IP address is exact 192.168.1.10, which by 'router-ips' is considered an router IP, and 'is-reachable' thought the router IPs are 'unreachable' and returned a false to the caller, thus, my host is determined to be 'not reachable' while it's perfectly reachable.

Relevant code here:
https://github.com/sindresorhus/is-reachable/blob/main/index.js#L55

I'm not sure I understand the rationale behind the decision, maybe logically 'is-reachable' should only check the 'reachability', whether it's an internal router IP or a normal external IP should be irrelevant, or am I missing something here?

[sindresorhus]

// @silverwind

[silverwind]

It's a failed heuristic. The idea behind router-ips is to detect cases where routers intercept HTTP requests, like captive portals for example. I think we better remove most IPv4's in router-ips that are not 1 in the last octet.

Or even simpler, remove the router-ips dependency altogether in this module, as it's probably causing more issues than it solves.

[silverwind]

I guess I would be in favor of just removing router-ips here. Guess it's better to just suceed a reachability check if there is something on the remote end, even if it's a captive portal. I recommend bumping semver major for this change.

[ccll]

That’s great, thanks for the explaination and effort here. ❤️