{"id":6695,"date":"2019-11-06T11:59:01","date_gmt":"2019-11-06T11:59:01","guid":{"rendered":"https:\/\/gtechbooster.com\/?p=6695"},"modified":"2023-06-21T11:42:52","modified_gmt":"2023-06-21T11:42:52","slug":"ipv4-addresses-running-out","status":"publish","type":"post","link":"https:\/\/gtechbooster.com\/ipv4-addresses-running-out\/","title":{"rendered":"IPv4 Addresses running out"},"content":{"rendered":"\n

RIPE, which manages IP address distribution for a large part of the world, has updated its address space chart showing how many addresses are left. It is clear that unless something unexpected happens RIPE will run out of addresses in November. <\/p>\n\n\n\n

\"\"<\/a>
RIPE NCC IPv4 Pool Stats<\/figcaption><\/figure><\/div>\n\n\n\n
\"\"<\/a><\/div><\/div>

The IPv4 crisis has been going on so long now that it is tempting to say “crisis, what crisis!” There are about 4 billion IP addresses and now we are down to just 1 million. That sounds a lot, but users tend to want blocks of many addresses and 1 million for the foreseeable future isn’t enough. There is already, as of October 2nd, a limit to only \/22 blocks (1024 addresses) being allocated and this will fall to \/24 blocks (256 addresses) once the well dries up. At this point applicants will form an orderly queue and wait for IP address allocations to be returned to RIPE. Only applicants who have never received an IP allocation will be added to the queue.<\/p>\n\n\n\n

What effect will this have?<\/h2>\n\n\n\n

In most cases end users don’t apply for IP addresses directly – they \nget them from their ISP or from their cloud or server provider. So for \nmost users it isn’t a matter of when RIPE runs out, more what happens \nwhen their ISP or server provider runs out. A recent survey by RIPE \nsuggests that about a third ranked IPv4 run-out as a top three challenge\n and more than half said that they needed more IPv4 addresses in the \nnext 2 to 3 years.<\/p>\n\n\n\n

Of course, like any resource run-out, estimates of doom and gloom are\n generally too pessimistic. Clever thinking makes the best of the \nsituation. IPv4 has been hugely extended by the use of dynamic address \nallocation and Carrier Grade NAT. ISPs can now operate their service \nwith fewer IP addresses. Companies which have stockpiled IPs are now \nable to make a profit on switching to NAT and selling their surplus IP \naddresses. It seems unlikely that November 2019 will go down in history \nas the day the IPv4 well ran dry, but things are going to get much more \ndifficult.<\/p>

\"\"<\/a><\/div><\/div>\n\n\n\n

When you think that IPv4 has 4 billion addresses and only servers \nneed a fixed IP, it doesn’t seem reasonable that we have used them all \nup. Perhaps it is time to force companies to relinquish IP addresses \nthat are needlessly allocated to servers that could just as easily be \nport-mapped. Can we do this automatically? Is it possible to scan IP \naddresses and make a list of uses that could be converted to NAT and \nport-mapping without a loss of functionality? We have spent a lot of \neffort getting everyone to upgrade to HTTPS, and this particular upgrade\n is arguably more important.<\/p>\n\n\n\n

We have been profligate with a resource that when first introduced \nseems so limitless that we could allocate huge blocks of addresses and \nlet anyone use them as they wanted to without any penalty.<\/p>\n\n\n\n

Surely the answer is IPv6, which really is a limitless address space? Yes, but changing over seems harder than anyone ever imagined.<\/p>\n\n\n\n

More Information<\/h2>\n\n\n\n