Neowin's plan to go both IPv4 and IPv6 in the future?


Recommended Posts

we would have to give each person on earth ~2 billion /64 segments in order to run out (assuming 7 billion people on earth).

 

Your talking the WHOLE space -- but that is not how it works..  We don't have the whole space to work with - you have what really to work with the 2000::/3 which is the global unicast space.. All the other space is assigned or reserved..

 

I get you -- its a LARGE number, but its still a finite number - if you don't pay attention you get yourself in trouble.. 

 

They talk about IPv4 being full - BS its not full, there are atleast 15 /8's that are "reserved for future use"  240-254, That's 250 million some addresses that could be used.  We all know ipv6 space is large, large does mean infinite ;)

Link to comment
Share on other sites

 

Your talking the WHOLE space -- but that is not how it works..  We don't have the whole space to work with - you have what really to work with the 2000::/3 which is the global unicast space.. All the other space is assigned or reserved..

 

I get you -- its a LARGE number, but its still a finite number - if you don't pay attention you get yourself in trouble.. 

 

They talk about IPv4 being full - BS its not full, there are atleast 15 /8's that are "reserved for future use"  240-254, That's 250 million some addresses that could be used.  We all know ipv6 space is large, large does mean infinite ;)

That only knocks off 3 bits, leaving a 2^125 bit sub-space to allocate from. If I redo the math from before: you are left with 2^61 /64 segments. So that means you'd have to give each person in existence 300 million /64 segments to exhaust the space. ~1/8th of what I said before. Or to say that another way, the global unicast segment is ~1/8th of the possible IPv6 addresses (2^3=8). Even if you dole out /64 segments for everyone, it is like having the equivalent of ~500 million IPv4 address spaces. Then if you manage to exhaust all of those addresses? You increment a bit (and use currently reserved space) and you have the equivalent of another ~500 million IPv4 address spaces if you decide to dole out /64 segments again. If you did that 8 times, you'd end up with the math I had in my previous post.

 

Finite in this sense may not be infinite, but it is a dontcare case for practical terms because it isn't something that will be an issue for millennia unless the rate of exhaustion becomes millions of times higher. Just to be clear about that bit of math, take something like 20 years to exhaust IPv4 and multiply that by 500 million then round down by orders and orders of magnitude to get millennia.  :laugh: It is hand wavy, but I'm rounding down to favor exhausting it much quicker.  ;)

Link to comment
Share on other sites

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.