IPv4's class E network ( contains 268 million addresses. Despite the advertisements for IPv6, claiming we have ran out of address space, this block ironically still claims to be "Reserved for future use". Why hasn't this block been freed up yet?

Of course, IPv6 should be promoted instead of freeing up more IPv4 addresses, but we've seen the address shortage coming for years. There has even been a time they weren't sure there was enough time to develop IPv6 before we would run out of addresses. Why didn't they free up this block already?

And is there any chance these addresses will be used in the future, like when IPv6 is fairly widely implemented but we still need IPv4 for backwards compatibility? It will be phased out regardless, but then ISPs don't have to employ NAT for IPv4 compatibility.

  • In hindsight, that address range should have been released for mobile IPv4 NAT. That would make it a lot easier for big Asian mobile companies to run NAT services for their ~1 billion users. – MSalters Aug 20 '12 at 10:36
  • @MSalters Hmm but it's up to the owner of the IP to decide whether or not to use NAT on it, right? They wouldn't have needed to be dedicated to this, if I understand correctly. Might have been used for both then. – Luc Aug 20 '12 at 20:52
  • the range never had an owner, but was reserved. "Use exclusively for NAT" could have been made a pre-condition when handing out allocations from that range. – MSalters Aug 21 '12 at 7:01


Networking stack implementations are written to support the RFC, and hence will not sent or accept packets from IPs that are reserved for future use. Besides the networking stack, services have to support them as well; DHCP has to be able to distribute them, DNS has to be able to store them, the software at IANA and your ISP must be able to actually support the creation and usage of that block.

By the time we all support this "class E" block, we will have made big progress switching to IPv6 so it'll no longer be worth it. Developers, ISPs and consumers better invest in switching to IPv6 instead...

However, many TCP/IP stacks, such as the one in Windows, do not accept addresses from class E space and will not even communicate with correspondents holding those addresses. It is probably too late now to change this behavior on the installed base before the address space would be needed

Read more (under Address Reclamation a Solution?)

  • 5
    Thought of that, but rejected the idea as too weird. Why would hard- and software be built not to work with those IPs, especially if it's written very clearly that the function of these addresses may change in the future? – Luc Aug 20 '12 at 6:30
  • @Luc: Because it is reserved for future use it doesn't mean that you can implement any form of support for it; this becomes more clear if we look at other RFCs. In proctocols, for example, they use that terminology to indicate some bits which you should not use as they could be used later for possible extensions to the protocol. As thus, I would suppose it to function the same for the IP space. You said it yourself, it's written very clearly that the function of these addresses may change in the future. They could possible not use them as just IP addresses but perhaps as something special... – Tamara Wijsman Aug 20 '12 at 9:03
  • @Luc: It's easy to try... ping on Windows. – Tamara Wijsman Aug 20 '12 at 9:09
  • 2
    @Luc: The reason they most likely have not released that block was because the subnet mask was part of it at that time and they didn't really care about running out back then, kind of like the famous "640K ought to be enough for anybody". So, it ended up being adapted this way. It would be rare to come up with that idea back then and even if you did, "If it ain't broken, don't fix it" applies, because that would require a lot of hassle for no immediate benefit. And well, RFC don't work like "first use it for this, then for that" but rather as a non-changing protocol, thus IPv4 -> IPv6... – Tamara Wijsman Aug 20 '12 at 21:26
  • 1
    @Luc: To get an idea, IPv4 (1979) got documented only 10 years after ARPANET (1969) went into operations; IPv6 (1998) got documented already 14 years ago, at the time IPv4 was about halfway exhausted. If they would have wanted to support the "class E" block they would have done so at that point, but why enable a small portion if you can invest time in a way bigger portion instead? If they have chosen to only support "class E" and delay IPv6 then we would actually risk exhaustion with no alternatives. In between IPv4 and IPv6 (1980 - 1990) there was probably no sign of exhaustion yet... – Tamara Wijsman Aug 20 '12 at 21:42

The problem is that this block is blacklisted in many operating systems. So they won't accept it as their own address and they won't connect to servers with such an address.

The other reason is that using class E would only extend the lifetime of IPv4 a year or so. Upgrading all operating systems for a year of extra delay with deploying IPv6 is not worth it

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.