hnetd
hnetd copied to clipboard
HomeNet-CP implementation (WIP)
If you disconnect the uplink on which a prefix was dynamically learned from DHCPv6 IA_PD, Hnet will immediately remove all routes pointing to assigned link /64s in that prefix. This...
ideas: - use scapy, so can pretend to be multiple devices? (with same MAC, but hopefully implementations don't do anything too low-level) - try to cover MUST/SHOULD/MAYs in the spec,...
delegated prefix: "prefix": "2000:dead:bee0:4::\/62", assignments: "prefix": "2000:dead:bee0:4:f::\/80", "prefix": "2000:dead:bee0:4:4::\/80", "prefix": "2000:dead:bee0:6:33::\/80", "prefix": "2000:dead:bee0:7:3f::\/80", "prefix": "2000:dead:bee0:5:9::\/80", "prefix": "2000:dead:bee0:4:10::\/80", Despite shortage and all, I would have expected at least some /64s :)
Valgrind shows various errors in test_pa\* (and also non-test PA code running vanilla hnetd). It would be nice if they were fixed at some point ;) (checks in test_pa_core test...
See subject. I think we had this same issue in autumn on OpenWrt, fixed it, but it is still on generic.
For footprint purposes, it would be 'nice to have'. However, how complicated that is in practice is hard to say. Currently only available in a 'preview' release so not urgent....
I wonder if we should consider merging dncp_link and iface somehow. (and whatever lives on the PA side of the fence.) Currently it is rather .. inefficient to have duplicate...
I got a string of issues (Starting with I dont think the cable modem is routing the delegation, but I am working on that separately), watching hnetd go... What I...