Ixp Report 2020
Ixp Report 2020
2020 Report
Contents
1. Introduction................................................................................................................................................................. 4
1.1 Foreword....................................................................................................................................................................... 4
1.2 Notes on this report ..................................................................................................................................................... 4
1.3 Internet Exchange Point (IXP) ...................................................................................................................................... 4
2. IXPs in Europe............................................................................................................................................................. 5
2.1 Number of operational IXPs in Europe......................................................................................................................... 5
2.2 Number of connected parties globally by IXP.............................................................................................................. 6
2.3 Number of connected parties in Europe by IXP........................................................................................................... 7
2.4 Number of connected parties globally by region .............................................................................................................. 8
6. Further Information.................................................................................................................................................. 20
The European Internet Exchange Association was established on June 28, 2001, in Amsterdam, The Netherlands. Since
this date, the European Internet Exchange Association (Euro-IX) has operated under the responsibility of an Association
according to Dutch law. Euro-IX has been set-up as a non-profit association.
Euro-IX brings together the Intenet community with the aim to share ideas and experiences and coordinate technical
standards, develop common procedures and publish statistics and other useful information. To compile this report on
IXPs, we used data from the Internet Exchange Point Database (IXPDB) - https://ixpdb.euro-ix.net/en/
The IXPDB is an authoritative, comprehensive, public source of data related to IXPs. It collects data directly from
IXPs through a recurring automated process. It also integrates data from third-party sources in order to provide a
comprehensive and corroborated view of the global interconnection landscape. For example, it highlights networks
that have not reported they are in PeeringDB and shows which IXPs are MANRS compliant. The combined data can be
viewed, analyzed, and exported through this web-based interface and an API.
In 2020, Euro-IX counted 72 Members and 11 Patrons from the IXP-related equipment-vendor and colocation sectors
of the community.
The 72 member IXPs are from 50 countries and the affiliations are as follows:
Euro-IX Region
NA Region
APIX Region
LAC-IX Region
Af-IX Region
1.1 Foreword
This report has been compiled by the European Internet Exchange Association (Euro-IX) to give an overview of the
current status of IXP deployment in Europe including:
- The number of Internet Exchange Points (IXPs) currently operating in Europe
- Related statistics and trends that are appearing in the European IXP market and a general global view
- The evolution over the last couple of years
Euro-IX has accepted the Internet Exchange Federations (IX-F) definition of an IXP being;
“A network facility that enables the interconnection and exchange of Internet traffic between more than two
independent Autonomous Systems. An IXP provides interconnection only for Autonomous Systems. An IXP does
not require the Internet traffic passing between any pair of participating Autonomous Systems to pass through any
third Autonomous System, nor does it alter or otherwise interfere with such traffic.”
The number of operational IXPs in Europe has increased by 87.5% over the past 10 years, going from 136 in
2010, to 255 in 2020. 2020 saw significant growth for IXPs, partially due to the number of IXPs added to the
IXPDB, indicating better reporting of active IXPs.
Note: The growth of 52 IXPs in 2020 includes both newly established IXPs but also IXPs that were only
identified by Euro-IX in 2020 (Source: IXPDB)
Graph 2.1.1
2010 18 4 14 136
2011 8 0 8 144
2012 2 0 2 146
2013 22 2 20 166
2014 37 16 21 187
2015 17 11 6 193
2016 44 12 32 225
2017 5 2 3 198
2018 7 2 5 203
2019 21 2 19 224
2020 59 7 52 255
Table 2.1.1
» The data below show the top 10 IXPs by connected parties globally. More complete data can
be found in the IXPDB.
2300
1800
1300
Number of ASNs
800
300
-200 IX.br Sao NAPAfrica IX IX.br Rio de IX.br Fortal eza SIX Seattle DE-CIX IX-SYD DE-CIX NYC IX.br Porto TorIX
Paulo Johannesburg Janeiro Mumbai (Sydney IX) Alegre
Graph 2.2.1
ASNs Unique ASNs
This data shows the total listed ASNs, and unique ASNs globally. The highest participants’ distribution is from
IX.br - São Paulo, Brazil, with 2203 ASNs, followed by NAPAfrica with 443 ASNs connected. If you see that an
IXP is missing, please speak to us or the IXP to get their data included.
» The data below shows the top 10 IXPs by connected parties in Europe. More complete data
can be found in the IXPDB.
1200
1000
800
Number of ASNs
600
400
200
0
DE-CIX AMS-IX LINX LON1 EPIX EPIX Warsaw MSK-IX NL-ix Piter-IX KleyReX Equinix Paris
Frankfurt
Graph 2.3.1
ASNs Unique ASNs
The data here shows the total listed ASNs, and unique ASNs in the Euro-IX region at the end of 2020. We see DE-
CIX Frankfurt leading the top ten with 1043 ASNs, followed by AMS-IX with 870 ASNs. However, it’s interesting to
see the IXP with the most unique ASNs is EPIX based in Poland, followed by MSK-IX and Piter-IX, both based in
Russia.
Table 2.3.1
» Euro-IX Region
30
3- 1
0 20 1 - 5 0 - 1 00 - 200 - 500 00
11- 2 51 - 10
101 201 50 1 Graph 2.3.1
Number of ASNs
» Af-IX Region
10
9
8
• In Africa, we see the largest majority of
7
Number of IXPs
Number of ASNs
28
24
20
Number of IXPs
Number of ASNs
» LAC-IX Region
22
20
18
16
14
12
Number of IXPs
0
0
00
0
-2
-5
Graph 2.3.4
20
50
3-
10
-1
11
21
1-
1-
51
1-
10
20
50
Number of ASNs
The average number of IXPs we collected data from was 56. These include Euro-IX members and “sub-members”, i.e.
other IXPs or PoPs operated by our members in Europe, such as LINX Manchester, Netnod Gothenburg etc.
Note: the aggregated peak traffic volume is determined by collecting publicly available traffic statistics from Euro-IX
Member IXPs. The data was captured during 12-month periods, from January to December each year, and varied from
one month to the other depending on the data available. Please refer to section 1.2 for details on how traffic data is
collected.
251
151
101
51
1
07 08 09 10 11 12 13 14 15 16 18 19 20
20 20 20 20 20 20 20 20 20 20 20 20 20
Graph 3.1.1
Aggregated peak IXP traffic within the Euro-IX Membership [in Gbps]
55,500
50,000
44,500
Number of IXPs
39,000
33,500
28,000
22,500
17,000
11,500
6,000
500
07 08 09 10 11 12 13 14 15 16 18 19 20
20 20 20 20 20 20 20 20 20 20 20 20 20 Graph 3.1.2
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Known operating IXPs in Europe 136 144 146 166 188 193 225 198 203 224 255
Table 3.1.1
These charts show general trends based on reported data. Not all IXPs report data or use the same measurement
intervals. These charts should be used to understand trends in data flows and not to place boundaries on total traffic
passing through IXP infrastructure.
100,000
95,000
90,000
85,000
Gbps
80,000
75,000
70,000
65,000
0 0 0 0 0 0 0 0 0 0
-20 b-2 r -2 r -2 y-2 n-2 l-2 g-2 p-2 t -2
0
v-2 c-2
Jan Fe Ma Ap Ma Ju Ju Au Se Oc No De
Graph 3.2.1
The average peak per IXP increased from 66,368 Gbps in January to 97,875 in December, with many peaks and
troughs, which can be attributed to the “lockdown” and “relaxing of lockdown” measures around the world.
» Number of IXPs monitored worldwide » Average Peak traffic per IXP (in Gbps)
700
155
650
IXPs
150
600
145
Gbps
550
140
135
500
130 450
-20 0 0
r -2
0 0 20 0 0 0 20 0 0 20 20 r -2
0
r -2
0 0 20 l-2
0
g-2
0 20 t -2
0
v-2
0
c-2
0
Jan b-2 r -2 y-2 Ju n- Ju
l-2 g-2 p-2 Oct - v-2 c-2 n- b- y-2 Ju n- Ju p- Oc
Fe Ma Ap Ma Au Se No De Ja Fe Ma Ap Ma Au Se No De
The number of IXPs monitored fluctuated throughout the year, starting with 147 in January, decreasing to 137 in
September, before going back up to 146 in December. The drop in data collected in September, can partially be
explained by the temporary decrease of IXPs monitored (with some IXP data unavailable at the time).
The average peak traffic per IXP increased with 49%, going from 451 Gbps to 670 Gbps. The traffic peaks in this
region ranged from 22 Mbps and 7,800 Gbps in January and from 25 Mbps in January to 10,181 Tbps in December.
The averages shown above are therefore not representative for most of the Exchanges and the IXP community
remains very diverse.
63,000
58,000
53,000
Gbps
48,000
43,000
20 0 0
n- b-
20 r -2
0
r -20 y-2 n-2 l-2
0
g-2
0
p-
20 t -2
0
v-2
0
c-2
0
Ja Fe Ma Ap Ma Ju Ju Au Se Oc No De
Graph 3.2.4
In the Euro-IX region, the average peak per IXP increased from 44,912 Gbps in January to 64,737 Gbps in December.
The peaks and troughs generally follow the same pattern as that in the global traffic, except traffic seemed to hold
steady nearing the end of 2020.
» Number of IXPs monitored in the Euro-IX region » Average Peak traffic per IXP (in Gbps)
930
73
880
830
70
Gbps
780
IXPs
730
67
680
64 630
0 0 0 0 0 0 0 0 0 0 0 0 0
-20 b-2 r -2
0
r -2
0
y-2 n-2 l-2
0
g-2
0 0
p-2 Oct -
20
v-2
0
c-2
0 -20 b-2 r -2 r -2 y-2 n-2 l-2 g-2 p-2 t -2
0
v-2 c-2
Jan Fe Ma Ap Ma Ju Ju Au Se No De Jan Fe Ma Ap Ma Ju Ju Au Se Oc No De
The number of IXPs monitored throughout the year varied between 72 in January, slightly decreasing to 69 in December.
The average peak traffic per IXP increased with 50%, going from 624 Gbps to 938 Gbps. The traffic peaks in this region
ranged from 31 Mbps to 3,500 Tbps in January and 77 Mbps to 10,181 Tbps in December.
The 4 largest IXPs in terms of traffic in this region are DE-CIX in Frankfurt, AMS-IX in Amsterdam, LINX in London and
MSK-IX in Moscow, which are the only ones where the traffic peaks above 3 Tbps each month.
1,970
1,820
1,670
1,520
1,370
1,220
Gbps
1,070
920
770
620
470
320
170
-20 0 0 0 0 0 0 0 0 0 0 0
b-2 r -2 r -2 y-2 n-2 l-2 g-2 p-2 t -2 v-2 c-2
Jan Fe Ma Ap Ma Ju Ju Au Se Oc No De
Graph 3.2.7
In the Af-IX region, the average peak per IXP increased from 1,034 Gbps in January to 1,556 Gbps in December.
Although there was some fluctuation throughout the year, the traffic seemed to steadily increase from September to
the end of the year. The sudden drop in May, September and November can partially be explained by the temporary
decrease of IXPs monitored during those months (with some IXP data unavailable at the time).
» Number of IXPs monitored in the Af-IX region » Average Peak traffic per IXP (in Gbps)
700
155
650
150
600
145
Gbps
IXPs
140 550
135 500
130 450
20 20 r -2
0
r -2
0
y-2
0
n-2
0
l-2
0 0 20 t -2
0 0 0
20 20 0 0 0 0 0 0 20 0 0 0
n- b- Ju g-2 p- Oc v-2 c-2 n- b- r -2 r -2 y-2 n-2 l-2 g-2 p- t -2 v-2 c-2
Ja Fe Ma Ap Ma Ju Au Se No De Fe Ju Oc
Ja Ma Ap Ma Ju Au Se No De
The average peak traffic per IXP increased with 69%, going from 148 Gbps to 250 Gbps. The traffic peaks in this
region ranged from 3.7 Gbps and 978 Gbps in January and from 25 Mbps in January to 1,556 Gbps in December.
The largest IXP in terms of traffic is NapAfrica (Johannesburg, Durban and Cape Town), now peaking above 1 Tbps
each month.
10,000
9,500
9,000
8,500
8,000
Gbps
7,500
7,000
6,500
6,000
5,500
5,000
0 0 0 0 0 0 0
-20 b-2 r -2 r -2
0
y-2 n-2 l-2
0
g-2 p-2 t -2
0
v-2
0
c-2
Jan Fe Ma Ap Ma Ju Ju Au Se Oc No De
Graph 3.2.10
In the APIX region, the average peak per IXP increased from 6,870 Gbps in January to 9,963 Gbps in December.
These numbers are based on data collected from 19 IXPs on average, over public peering LAN. The sudden drop in
September can be explained by the temporary decrease of IXPs monitored during that month (with some IXP data
unavailable at the time).
» Number of IXPs monitored in the APIX » Average Peak traffic per IXP (in Gbps)
region
20 590
19 540
18 490
IXPs
Gbps
17 440
16 390
15 340
-20 eb-20 ar -20 pr -20 ay-2
0 0 0 0 0 20 0 0
n-2 l-2 g-2 p-2 Oct - v-2 c-2 -20 0 0
r -2
0
y-2
0
n-2
0 0 0 0 20 0 0
Jan F A Ju Ju Au Se No De Jan b-2 r -2 Ju
l-2 g-2 p-2 Oct - v-2 c-2
Ma Ap Ma
M M Ju
Fe Au Se No De
The number of IXPs monitored throughout the year varied between 19 in January, slightly decreasing to 17 in December.
The average peak traffic per IXP increased with 62%, going from 362 Gbps to 586 Gbps. The traffic peaks in this region
ranged from 6.5 Gbps to 1,996 Gbps in January and 35 Gbps to 2,770 Gbps in December.
The 3 largest IXPs in terms of traffic in this region are JPNAP, JPIX and Hong Kong IX.
14,000
13,000
12,000
11,000
10,000
9,000
IXPs
8,000
7,000
6,000
5,000
0 0 0
-20 b-2 r -2
0
r -2
0
y-2 n-2 l-2
0
g-2
0
p-2
0
t -2
0
v-2
0
c-2
0
Jan Fe Ma Ap Ma Ju Ju Au Se Oc No De
Graph 3.2.13
In the LAC-IX region, the average peak per IXP increased from 8,807 Gbps in January to 13,993 Gbps in December.
Unlike in the other regions, there was no sudden decrease of traffic occurring in the year. The traffic has been
growing steadily with slight drops in April and August.
» Number of IXPs monitored in the LAC-IX region » Average Peak traffic per IXP (in Gbps)
36 450
35 400
34 350
33 300
Gbps
IXPs
32 250
31 200
30 150
20 20 0 0 0 0 0 0 20 0 0 0
20 20 r -2
0
r -2
0 0 20 l-2
0
g-2
0 20 t -2
0
v-2
0
c-2
0 n- b- r -2 r -2 y-2 n-2 l-2 g-2 p- t -2 v-2 c-2
n- b- y-2 Ju n- Ju p- Oc Ja Fe Ma Ap Ma Ju Ju Au Se Oc No De
Ja Fe Ma Ap Ma Au Se No De
The number of IXPs monitored throughout the year varied between 35 in January, slightly decreasing to 32 in May
and going back up to 35 in December. The average peak traffic per IXP increased with 98%, going from 252 Gbps to
400 Gbps. The traffic peaks in this region ranged from 31 Mbps to 3,500 Tbps in January and 1,08 Gbps to 10,065
Gbps in December.
The largest IXP in terms of traffic in this region is IX.br (São Paulo), where the traffic peaks above 7 Tbps each month.
IX.br (São Paulo) and DE-CIX Frankfurt have the most data throughput worldwide.
100,000
95,000
90,000
85,000
Gbps
80,000
75,000
70,000
65,000
20 20 0 -20 0 0 0 0 20 0 0 0
n- b- r -2 r y-2 n-2 Ju
l-2 g-2 p- t -2 v-2 c-2
Ja Fe Ma Ap Ma Ju Au Se Oc No De
Graph 3.2.16
In the North America (NA) region, the average peak per IXP increased from 4,745 Gbps in January to 7,182 in
December. The NA region experienced fluctuations similar to the other regions, with traffic steadily increasing from
September to the end of the year.
» Number of IXPs monitored in the NA region » Average Peak traffic per IXP (in Gbps)
17
420
16 400
380
15
Gbps
360
IXPs
340
14
320
13 300
20 0 0 0 0 0
n-
20 b- r -2
0
r -2
0
y-2 n-2 l-2
0
g-2
0
p-
20 t -2
0
v-2
0
c-2
0 -20 eb-2 r -2
0
r -2
0
y-2 n-2 l-2
0
g-2
0 0
p-2 Oct -
20
v-2
0
c-2
0
Ja Fe Ma Ap Ma Ju Ju Au Se Oc No De Jan F Ma Ap Ma Ju Ju Au Se No De
The number of IXPs we collect data from went from 14 in January to 17 in December. The average peak traffic per
IXP increased with 24%, going from 339 Gbps to 422 Gbps. The traffic peaks in this region ranged from 43 Mbps to
1,300 Tbps in January and 1.45 Gbps to 1,810 Tbps in December.
The largest IXPs in this region are SIX in Seattle, Telx Atlanta (both showing a traffic peaking above 1 Tbps each
month), and NYIIX in New York and Torix in Toronto, with a steadily increasing monthly peak.
» The tables show the top 10 IXPs and countries in the world with the highest traffic. More complete data
can be found in the IXPDB.
» IXPs with the most traffic in Europe » Countries with the most IXP traffic in Europe
Maximum in
IXP
Gbps
1 IX.br 13,047
2 JPNAP Tokyo I 2,800
3 IX.br Rio de Janeiro 1,890
4 SIX - Seattle 1,850
5 HKIX 1,840
6 NAPAfrica 1,556
7 JPIX Tokyo 1,530
8 TLEX Atlanta 1,290
9 TORIX 1,166
10 IX.br Fortaleza 1,021
Table 3.3.3
35
30
Cisco
25
Arista
20 Juniper Networks
15 Extreme Networks
Cumulus Networks
10
0
Cisco Arista Juniper Extreme Cumulus
Graph 4.1.1
Networks Networks Networks
6%
3%
2% 14%
6%
Cisco 34%
6%
Arista
34%
Juniper Networks
7% 16%
Extreme Networks
Cumulus Networks20%
21%
30%
Chart 4.1.1
250
Extreme
200 Arista
Cisco
150
Juniper Networks
Huawei
100
Cumulus Networks
50
0
Extreme Arista Cisco Juniper Networks Huawei Cumulus Graph 4.2.1
Networks
4%
21%
17% 9%
Extreme
10% 31%
10%
Arista
Cisco 31%
6%
2%
Huawei 2%
2% 2%
Chart 4.2.1
6.1 Resources
IXPDB - The data in the IXPDB is automatically updated from JSON feeds supplied directly by the IXPs. The
data, in addition to member lists and location information, also includes details of hardware used at IXPs.
We thank all sponsors of the IXPDB, Euro-IX Members and Patrons. We additionally thank all European IXPs,
and especially those that provide publicly available information.
6.2 Contact
We welcome all forms of feedback and suggestions concerning this report and will do our best to answer any
further requests for information.
# OF
ECONOMY CITY FULL NAME OF IXP ACRONYM
IXPS
Albania 1
Tirana Albanian Neutral Internet Exchange ANIX*
Armenia 1
Yerevan Armenia Internet Exchange ARMIX*
Austria 5
Graz Grazer Internet Exchange GraX
Klagenfurt Alpes Adria Internet Exchange AAIX
Vienna Community Internet Exchange - Vienna Community-IX Vienna
Vienna Equinix Internet Exchange - Vienna Equinix Vienna**
Vienna Vienna Internet eXchange VIX*
Belarus 1
Minsk Belarus National Internet Exchange Point BY-IX
Belgium 2
Brussels Belgian National Internet Exchange BNIX*
Waasmunster Belgium Internet Exchange BelgiumIX
Bosnia and Herzegovina 1
Sarajevo Bosnia and Herzegovina Internet Exchange BHNIX*
Bulgaria 5
Sofia Balkan Internet Exchange Balkan-IX
Sofia Bulgarian Internet Exchange BIX.BG*
Sofia Sofia Internet Exchange IX-SOF
Sofia NetIX NetIX*
Sofia T-CIX T-CIX
Croatia 1
Zagreb Croatian Internet eXchange CIX*
Cyprus 1
Nicosia Cyprus Internet Exchange CyIX
Czech Republic 3
Brno Brno Internet Exchange Point BR-IX
Prague Neutral Internet eXchange of Czech Republic NIX.CZ*
Prague Peering.cz Peering.cz*
Denmark 3
Copenhagen Danish Internet eXchange point DIX*
Copenhagen Netnod - Copenhagen Netnod - Copenhagen**
Copenhagen Stockholm Internet Exchange - Copenhagen STHIX Copenhagen**
Estonia 4
Riga PITER-IX PITER-IX Riga
Tallinn Tallinn Governmental Internet Exchange RTIX
Tallinn Tallinn Internet eXchange TIX Tallinn
Tallinn Tallinn Internet Exchange TLLIX
Mutually Agreed Norms for Routing Security (MANRS) is a global initiative, supported by the Internet Society, that
provides crucial fixes to reduce the most common routing threats.
In 2020, there were 109 IXPs who were MANRS in the IXPDB.
600
500
400
300
IXPB Data
200
Total IXPs 651
100 New IXPs added in 2020 59
0
MANRS Compliant IXPs 109
Total IXPs New IXPs MANRS
compliant
Euro-IX Region
NA Region = 224 API’s
= 123 API’s
APIX Region
= 134 API’s
LAC-IX Region
= 96 API’s
Af-IX Region = 39
API’s