[HN Gopher] The IPv4 address swamp: The new normal
___________________________________________________________________
The IPv4 address swamp: The new normal
Author : speckx
Score : 22 points
Date : 2025-12-23 16:11 UTC (1 days ago)
(HTM) web link (blog.apnic.net)
(TXT) w3m dump (blog.apnic.net)
| AtlasBarfed wrote:
| Setting aside the address scarcity issue, how is IPv6 going to
| simplify the routing table? If anything, it would just be an
| explosion of the number of addresses?
|
| I mean a million is objectively a large number if it's all on
| paper, but to me, that's not a particularly large data set for
| talking about the entire freaking internet.
|
| And how cheap of a SOC can handle that in memory? A better
| question might be to even make a system on a chip that couldn't
| handle that memory?
| _bernd wrote:
| The issue is; in the default free zone, every peer which gives
| you a full table, gives you 1 million routes. Core
| infrastructure is not getting refreshed every 5 year, I have
| heard so...
| slyall wrote:
| There theory might be that an organisation would end up
| advertising a single prefix, rather than whatever they have now
| (say 40 networks with various prefixes).
| orangeboats wrote:
| A single /32 IPv6 prefix is actually easier on the router
| (computational and memory wise) than a dozen /24 IPv4 prefixes.
| 486sx33 wrote:
| Huh A single prefix is easier on the router than a dozen.. I
| should hope so? Isn't this kind of like saying the grade 1
| math test is easier than the grade 12 math test ?
| teraflop wrote:
| The thing is that the abundance of IPv6 addresses _enables_
| fewer prefixes to be used, by allowing addresses to be
| allocated in much larger chunks.
|
| For instance, Comcast (AS 7922) owns about 2^26 IPv4
| addresses, distributed across 149 different prefixes.
| Almost all of these prefixes are non-contiguous with each
| other, so they each require separate routing table entries.
| Comcast can't consolidate those routes without swapping IP
| address blocks with other networks, and it can't grow its
| address space without acquiring new small blocks. (Since no
| more large blocks are available, as this article
| discusses.)
|
| In contrast, Comcast owns about 2^109 IPv6 addresses, which
| are covered by just 5 prefixes (two big ones of 2^108 each,
| and three smaller ones). It can freely subdivide its own
| networks within those prefixes, without ever running out of
| addresses, and without having to announce new routes.
| rcxdude wrote:
| It's not just any memory. When it comes to core infrastructure
| routers those routes need to fit into specialized and expensive
| CAM (Content Addressable Memory) to do the lookups in hardware.
| And on every single one.
| tptacek wrote:
| Right but that's still not really answering his question.
| Sure, the constant factor is higher for router TCAM memory.
| Still: you can sum this post up as "in the late 1990s, tier-1
| carriers filtered advertisements for all but the 'swamp'
| range down to /19s or smaller prefixes; now everything is the
| 'swamp'". Why is that?
| jandrese wrote:
| Because IPv4 address scarcity means small blocks get sold
| as they are available to people in completely different
| parts of the Internet. With IPv6 the address space is so
| large that they can easily keep the blocks in one piece.
| tptacek wrote:
| No, obviously, I get that (we buy a lot of IPv4 space ---
| and I'm actually happier with the current regime than I
| was with the "supplicate to ARIN" regime). I'm just
| wondering what technologically happened to make universal
| /24 advertisements fine. I assume it's just that routers
| got better.
| toast0 wrote:
| The small ISP that serves my home has six IPv4 prefixes and one
| IPv6 prefix.
|
| My small hosting provider has I think 7 v4 prefixes, but could
| be one v6 prefix. Maybe not --- a lot of their /22s are
| advertised as four /24s to allow for a DDoS Mitigation provider
| to attract traffic when needed; but it'd probably still be
| fewer prefixes with v6.
|
| Not every ASN looks the same, but many of them would advertise
| a lot fewer prefixes if they could get contiguous addresses,
| but it's not possible/reasonable to get contiguous allocations
| for v4.
|
| Since the routing table is organized around prefixes, _if_
| there is complete migration, the routing table will probably be
| smaller.
| tptacek wrote:
| This post is kind of a weird promotion for NETSCOUT, written by
| an analyst on the Arbor ATLAS team (NETSCOUT owns Arbor now).
___________________________________________________________________
(page generated 2025-12-24 23:00 UTC)