IPv4

Addresses Exhaus0on,
not only when is cri0cal
Arturo Servn
LACNIC

All started
When in LACNIC we realized that we will not
get more IPv4 address space from IANA
And we decided to publish how much IPv4
address space we had leI
Then, we were curious of when this space
was going to run out

The numbers

Available Today
Addresses Available Today

63161600

Available today /8

3.764724731

Reserve Last /10 (policy)

-0.25

Global Policy

Total Number of Addresses 4.5152 /8 = 75,744,512 addresses

200501
200505
200509
200601
200605
200609
200701
200705
200709
200801
200805
200809
200901
200905
200909
201001
201005
201009

Montlhy Address Alloca0ons

Total /8

0.6

0.5

0.4

0.3

0.2
Total /8

0.1

200711
200712
200801
200802
200803
200804
200805
200806
200807
200808
200809
200810
200811
200812
200901
200902
200903
200904
200905
200906
200907
200908
200909
200910
200911
200912
201001
201002
201003
201004
201005
201006
201007
201008
201009
201010
201011

Annualized
(36 months)

1.6

1.4

1.2

0.8

0.6

0.4

0.2

So, we know how much we have

and how much we use, so lets
make some numbers

Usage Projec0ons
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0

Polin. Mod.
Log. Mod.
Lin. Mod.

201403

201312

201309

201306

201303

201212

201209

201206

201203

201112

201109

201106

201103

201012

201009

Exp. Mod.

It appears that we will have addresses

for almost 4 more years.
But there is more to it than when

Availability vs. Accessibility

Diculty (and cost)
Of gedng IPv4
addresses

Availability of IPv4
Addresses space
t

Modeling Diculty
We model diculty as func0on of the
number of requirements to apply for new
address space and the amount of addresses
that a RIR member would be able to get per
year
Requirements are: More strict alloca0ons
policies
Among them allowing only one address
request of a specic size per year

Experimental Diculty
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0

Experimental
Diculty

Sep-14

May-14

Jan-14

Sep-13

May-13

Jan-13

Sep-12

May-12

Jan-12

Sep-11

May-11

Jan-11

IPv4 remaining
space

Sep-14

May-14

Jan-14

Sep-13

May-13

Jan-13

Sep-12

May-12

Jan-12

Sep-11

May-11

Jan-11

Experimental Diculty

5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
IPv4 + Dif

IPv4

Uncertainty Factor

Lineal Projec0on
Low uncertainty factor
High uncertainty factor

2013

What is uncertainty?
A new technology demanding connec0vity,
ergo an increase in IPv4 addresses requests
Mobile devices
Transi0on mechanisms to IPv6

Rush shopping

Jan-11
Apr-11
Jul-11
Oct-11
Jan-12
Apr-12
Jul-12
Oct-12
Jan-13
Apr-13
Jul-13
Oct-13
Jan-14
Apr-14
Jul-14
Oct-14

Experimental Uncertainty Factor

5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
IPv4

Disrup

IPv4+Disr

Jul-14

Apr-14

Jan-14

Oct-13

Jul-13

Apr-13

Jan-13

Oct-12

Jul-12

Apr-12

Jan-12

Oct-11

Jul-11

Apr-11

Jan-11

All together

5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
IPv4
All

As seen, projec0on may be

dicult, look

New Projec0on (Dec 2010)

It moved
right in
November

Lin. Model
Orig

Real + Lin
Model Data
Nov.
201009
201012
201103
201106
201109
201112
201203
201206
201209
201212
201303
201306
201309
201312
201403

5
4.5
4
It moved 3.5
leI in
3
October 2.5
2
1.5
1
0.5
0

Conclusions
Modeling the IPv4 address exhaus0on is not
easy, but that is not a surprise
Close monitoring is crucial if we want to
react to changes in the ecosystem
Modeling short periods of 0me seems to be
a sensible approach, but then RIRs need to
open up their stats about stock and
alloca0on

Conclusions (cont.)
Knowing the when is important, but
how the predic0on changes and why is
also cri0cal
Alloca0on policies can extend the life of the
RIR pool but at a cost of availability
New technologies and rush shopping may
signicantly aect the exhaus0on model

200501
200504
200507
200510
200601
200604
200607
200610
200701
200704
200707
200710
200801
200804
200807
200810
200901
200904
200907
200910
201001
201004
201007
201010
201101

Sleep 0ght

100000000
10000000

1000000

100000

APNIC

10000

RIPE

1000

ARIN

100

LACNIC

10

Afrinic

Thank you!
aservin@lacnic.net
Twiqer:@the_real_r2d2