> I can do a rough inventory of my IPv4 home network (120 devices) by looking at hostnames for DHCPv4 leases, no such thing exists for IPv6. You mean, you can

> I can do a rough inventory of my IPv4 home network (120 devices) by looking at hostnames for DHCPv4 leases, no such thing exists for IPv6. You mean, you can do a rough inventory of IPv4 devices that are well-behaved and want to be found. Unless you have enterprise-y network gear, which can do port isolation and preferably 802.1x auth, it's trivial for a host to poison your switch's ARP tables to check whether the address it wants to use it free, and then simply use that address without requesting it from the DHCP server[1]. The DHCP server will check that no hosts are using that address (usually via ICMP) before assigning any leases to it, but won't (at least for any common DHCP servers) register a lease for any host it discovers, so the rogue device doesn't even need to worry about changing addresses when a new host joins the network. At best, you might discover the presence of such a device by checking your DHCP server's logs. Since no common consumer-grade NAT implementations check whether the internal host has a valid lease before establishing NAT state, the rogue internal device is free to communicate with any external host. Even better, the rogue device can send packets with a from-address of other devices on your network, allowing it to punch arbitrary holes in your NAT. As other commenters mentioned, you can get a similar level network surveillance for IPv6 using mDNS, but at least IPv6 isn't lulling you into a false sense that you can see all the devices on your network in the lease table. [1] Actually, it doesn't even need to do that... it can simply start using an address and then change to another address if connectivity seems to be a problem.