|
Mid March a special plenary session of the Canadian standard committee isacc was convened in Ottawa to review the final report [PDF] of the Canadian IPv6 Task Group. It was unanimously approved and the essence of its 66 pages are seven recommendations for Government, Industry, Service and Content providers, and the regulator, CRTC to proceed with diligence, even some sense of urgency.
One paragraph provides an interesting new twist to the exhaustion debate: Is the Internet already becoming less reliable as a consequence? paragraph 3.1 of the report says:
‘Evidence has shown that most of the remaining IPv4 address space is already in use by organizations. Within the few remaining IPv4 address blocks available as of January 2010, 90% of that address space contains prefixes which have been identified as already in use by some organizations, resulting in decreased reliability. Therefore, the recipients of these prefixes, when allocated, will see unwanted traffic to their networks and many organizations will not be able to reach these recipients’ networks. In other words, the remaining address space will be less reliable to use than the IPv4 address space already in use. As the remaining address space approaches zero, it is likely that people will experience unreachability of sites and networks as well as more instability in IPv4 routing.’
Could we be using dirty prefixes without knowing it? What are the possible consequences? How to make sure we get clean addresses from our ISP? Marc Blanchet’s blog on the topic might provide some beginnings of an answer.
If you consider that the IPv4 address pool has only 24 prefixes (/8’s) left and that 22 of them are dirty, means only two really clean ones are left! The dirtiest neighbourhoods are 1.0.0.0/8, 2.0.0.0/8 and 100.0.0.0/8 , they’re apparently harbouring nests of address squatters.
An IPv4 dirtyness index might be a useful addition to the expiry counters. A healthier alternative however might be to start dipping in the IPv6 reservoir.
Sponsored byWhoisXML API
Sponsored byDNIB.com
Sponsored byIPv4.Global
Sponsored byVerisign
Sponsored byCSC
Sponsored byRadix
Sponsored byVerisign
While it’s important to be aware of this problem, it’s not news that it exists. As I pointed out last week ICANN has been writing and presenting on this problem for a few years now and commission earlier research into it. But perhaps more importantly, as George Michaelson and Geoff Huston reported from their research into the pre-existing, unauthorized usage of 1.0.0.0/8:
We should not assume that just because there is some unauthorized usage in unallocated address space in parts of a /8 that the whole block is unusable. There will undoubtedly be some problems with these prefixes. There always are problems when new address space is brought into official use but those problems get fixed and people move on.
Some of them have got surprisingly large sized allocations - a significant number of new allocations (including /15 and /16 netblocks) as well as hijacking and repurposing of legacy netblocks that belonged to a defunct organization). With v4 being in such short supply that needs to be addressed far more aggressively than it currently is. In any case, even if that IP space is reclaimed, getting it out of various blocklists, nullroutes etc will be an exercise that I wouldn't wish on my worst enemy