302-761-5782

About this page This page presents a dataset on traffic load balancing practices in the Internet, for both IPv4 and IPv6. We deployed our open source tool in several monitors around the world and conduct measurements towards different IP destinations. Our tool implements a modified version of the Multipath Detection Algorithm and is capable of identifying load balancing as well as enumerating packet's header fields used by routers for traffic load balancing. We identify per-packet load balancing and classify different traffic load balance practices, such as the use of the Flow Label header field in IPv6. This page brings a real-time summary and preliminary analysis of our ongoing measurements. We also make our dataset available for download and exploration through a web based tool.

Monitors

Last update on 2018-10-18 10:49:27 Z

Monitor Host Location Last update (UTC) Per hour Traces Dataset Status
hel-fi-ark Ark Espoo, Finland 2018-10-14 05:24:47 35.18 45215 meliority · Download Down
syd-au-ark Ark Sydney, Australia 2018-10-13 19:08:39 35.50 45266 Explore · (850) 670-6656 Down
cbg-uk-ark Ark Cambridge, United Kingdom 2018-10-15 16:38:55 34.20 45167 (276) 724-0474 · 8656371804 Down
zrh-ch-ark Ark Zurich, Switzerland 2018-10-13 03:44:07 35.76 45050 9374807875 · magnifice Down
bma-se-ark Ark Kista, Sweden 2018-10-10 04:27:42 24.92 29621 6465462004 · 5615711836 Down
per-au-ark Ark Perth, Australia 2018-10-18 10:47:42 26.15 36261 Explore · 5035494529 Down
yyz-ca-ark Ark Toronto, Canada 2018-10-13 03:42:46 35.92 45124 Explore · Download Down
nrt-jp-ark Ark Tokyo, Japan 2018-10-18 10:49:27 30.11 41757 Explore · 7724692263 Down
cjj-kr-ark Ark Daejeon, South Korea 2018-10-12 13:04:12 36.20 45070 Explore · (479) 521-2088 Down
dub-ie-ark Ark Dublin, Ireland 2018-10-13 23:00:00 35.61 45548 Explore · Download Down
tpe-tw-ark Ark Hsinchu, Taiwan 2018-09-16 21:57:11 31.07 19574 Explore · Download Down
yyz-ca-doc DigitalOcean Toronto, Canada 2018-09-09 15:02:49 103.97 47450 6463380678 · 2625838815 Down
blr-in-doc DigitalOcean Bangalore, India 2018-09-10 16:24:03 98.46 47433 Explore · Download Down
sfo-us-doc DigitalOcean San Francisco, CA 2018-09-08 14:27:42 110.05 47521 Explore · Download Down
ams-nl-doc DigitalOcean Amsterdam, Netherlands 2018-09-19 23:43:22 66.07 46584 Archy · Download Down
lga-us-doc DigitalOcean New York, NY 2018-09-08 14:01:19 109.57 47262 (662) 799-5429 · Download Down
sin-sg-doc DigitalOcean Singapore, Singapore 2018-09-10 07:43:59 99.62 47129 Explore · Download Down
fra-de-doc DigitalOcean Frankfurt, Germany 2018-09-09 09:14:52 105.22 47409 808-739-6917 · 7086531820 Down
atl-us-lin Linode Atlanta, GA 2018-09-21 08:43:55 63.49 46758 Explore · (778) 475-9242 Down
frm-us-lin Linode Fremont, CA 2018-09-17 11:01:06 72.68 46832 Explore · Download Down
dfw-us-lin Linode Dallas, TX 2018-09-14 01:59:14 83.46 47016 Explore · 3462785794 Down
ewr-us-lin Linode Newark, NJ 2018-09-10 19:34:37 96.82 46951 Explore · 822-935-8334 Down
hnd-jp-lin Linode Tokyo, Japan 2018-09-16 07:20:26 75.99 46865 (970) 626-1661 · Download Down
lcy-uk-lin Linode London, United Kingdom 2018-09-17 09:35:10 73.12 47012 Explore · Download Down
bhz-br-ufmg UFMG Belo Horizonte, Brazil 2018-09-06 11:30:57 82.26 31552 Explore · Download Down
syd-au-vul Vultr Sydney, Australia 2018-09-14 16:16:10 78.97 45611 Explore · Download Down
mia-us-vul Vultr Miami, FL 2018-09-27 01:40:42 53.23 46574 Explore · Download Down
sea-us-vul Vultr Seattle, WA 2018-09-12 14:48:47 89.13 47073 (513) 782-7598 · Download Down
lax-us-vul Vultr Los Angeles, CA 2018-09-13 13:19:57 85.97 47338 Explore · Download Down
ord-us-vul Vultr Chicago, IL 2018-09-11 07:05:31 95.26 47290 Explore · Download Down
cdg-fr-vul Vultr Paris, France 2018-09-11 18:47:22 92.64 47073 Explore · (669) 204-9323 Down

Analysis summary

Occurrence of load balancing

Figures 1 and 2 show the total number of load balancers per route identified in our dataset, for IPv4 and IPv6, respectively. We highlight the greater occurrence of load balancing in IPv4 routes compared to IPv6: ~20% of IPv4 routes traverse more than 10 load balancers, in contrast to only ~5% for IPv6 routes.

Figures 3 and 4 show the total number of diamonds per route identified in our dataset, for IPv4 and IPv6, respectively. Although the occurrence of diamonds is very similar for both protocols, we observe IPv4 routes traversing slightly more diamonds than IPv6 routes.

Load balancing classification

Field IPv6 IPv4
UDP TCP ICMP UDP TCP ICMP
Any 21522 (100%) 16730 (100%) 22390 (100%) 41917 (100%) 43154 (100%) 45792 (100%)
17344 (100%) 13591 (100%) 19438 (100%) 32794 (100%) 33702 (100%) 40674 (100%)
17119 (100%) 13373 (100%) 19253 (100%) 32123 (100%) 32925 (100%) 40284 (100%)
daddr+dport 11682 (54.28%) 8572 (51.24%) - 18431 (43.97%) 16864 (39.08%) -
11385 (65.64%) 8359 (61.50%) - 17375 (52.98%) 15734 (46.69%) -
11162 (65.20%) 8316 (62.18%) - 17930 (55.82%) 16339 (49.62%) -
daddr 4523 (21.02%) 4543 (27.15%) 16557 (73.95%) 15898 (37.93%) 18188 (42.15%) 38597 (84.29%)
3099 (17.87%) 3382 (24.88%) 16278 (83.74%) 12678 (38.66%) 14994 (44.49%) 37957 (93.32%)
2981 (17.41%) 3315 (24.79%) 15526 (80.64%) 12110 (37.70%) 14339 (43.55%) 37765 (93.75%)
daddr+dport+fl 2178 (10.12%) 1003 (6.00%) - - - -
1949 (11.24%) 921 (6.78%) - - - -
2134 (12.47%) 960 (7.18%) - - - -
dport 1357 (6.31%) 901 (5.39%) - 2427 (5.79%) 2293 (5.31%) -
120 (0.69%) 83 (0.61%) - 367 (1.12%) 353 (1.05%) -
54 (0.32%) 40 (0.30%) - 138 (0.43%) 130 (0.39%) -
daddr+dport+tc 384 (1.78%) 348 (2.08%) - - - -
279 (1.61%) 265 (1.95%) - - - -
345 (2.02%) 308 (2.30%) - - - -
daddr+fl 103 (0.48%) 104 (0.62%) 2377 (10.62%) - - -
46 (0.27%) 64 (0.47%) 2047 (10.53%) - - -
36 (0.21%) 45 (0.34%) 2240 (11.63%) - - -
daddr+dport+fl+tc 70 (0.33%) 57 (0.34%) - - - -
6 (0.03%) 7 (0.05%) - - - -
76 (0.44%) 71 (0.53%) - - - -
dport+fl 58 (0.27%) 42 (0.25%) - - - -
14 (0.08%) 8 (0.06%) - - - -
4 (0.02%) 5 (0.04%) - - - -
daddr+tc 57 (0.26%) 39 (0.23%) 346 (1.55%) - - -
7 (0.04%) 5 (0.04%) 167 (0.86%) - - -
9 (0.05%) 5 (0.04%) 248 (1.29%) - - -
per-packet 49 (0.23%) 34 (0.20%) 343 (1.53%) 88 (0.21%) 103 (0.24%) 147 (0.32%)
22 (0.13%) 16 (0.12%) 81 (0.42%) 48 (0.15%) 55 (0.16%) 67 (0.16%)
16 (0.09%) 13 (0.10%) 59 (0.31%) 32 (0.10%) 37 (0.11%) 45 (0.11%)
dport+tc 36 (0.17%) 44 (0.26%) - - - -
5 (0.03%) 6 (0.04%) - - - -
3 (0.02%) 5 (0.04%) - - - -
daddr+fl+tc 21 (0.10%) 14 (0.08%) 81 (0.36%) - - -
15 (0.09%) 12 (0.09%) 17 (0.09%) - - -
14 (0.08%) 11 (0.08%) 67 (0.35%) - - -
dport+fl+tc 10 (0.05%) 15 (0.09%) - - - -
4 (0.02%) 5 (0.04%) - - - -
2 (0.01%) 2 (0.01%) - - - -
fl 9 (0.04%) 13 (0.08%) 82 (0.37%) - - -
2 (0.01%) 4 (0.03%) 18 (0.09%) - - -
1 (0.01%) 1 (0.01%) 10 (0.05%) - - -
tc 8 (0.04%) 10 (0.06%) 45 (0.20%) - - -
3 (0.02%) 5 (0.04%) 4 (0.02%) - - -
3 (0.02%) 4 (0.03%) 4 (0.02%) - - -
fl+tc 4 (0.02%) 5 (0.03%) 4 (0.02%) - - -
- 1 (0.01%) - - - -
- - - - - -
chksum+daddr - - 203 (0.91%) - - 717 (1.57%)
- - 17 (0.09%) - - 324 (0.80%)
- - 109 (0.57%) - - 429 (1.06%)
chksum+daddr+tc - - 174 (0.78%) - - -
- - 132 (0.68%) - - -
- - 153 (0.79%) - - -
chksum+daddr+fl+tc - - 68 (0.30%) - - -
- - 4 (0.02%) - - -
- - 61 (0.32%) - - -
chksum+daddr+fl - - 59 (0.26%) - - -
- - 3 (0.02%) - - -
- - 40 (0.21%) - - -
chksum - - 27 (0.12%) - - 33 (0.07%)
- - 9 (0.05%) - - 6 (0.01%)
- - 11 (0.06%) - - 4 (0.01%)
chksum+tc - - 16 (0.07%) - - -
- - 5 (0.03%) - - -
- - 4 (0.02%) - - -
chksum+fl+tc - - 15 (0.07%) - - -
- - 4 (0.02%) - - -
- - 3 (0.02%) - - -
chksum+fl - - 2 (0.01%) - - -
- - - - - -
- - - - - -
chksum+daddr+tos - - - - - 642 (1.40%)
- - - - - 427 (1.05%)
- - - - - 626 (1.55%)
daddr+tos - - - 143 (0.34%) 129 (0.30%) 567 (1.24%)
- - - 31 (0.09%) 29 (0.09%) 265 (0.65%)
- - - 46 (0.14%) 37 (0.11%) 424 (1.05%)
chksum+tos - - - - - 61 (0.13%)
- - - - - 20 (0.05%)
- - - - - 15 (0.04%)
tos - - - 38 (0.09%) 51 (0.12%) 45 (0.10%)
- - - 20 (0.06%) 21 (0.06%) 14 (0.03%)
- - - 15 (0.05%) 22 (0.07%) 10 (0.02%)
daddr+dport+tos - - - 989 (2.36%) 1004 (2.33%) -
- - - 684 (2.09%) 683 (2.03%) -
- - - 977 (3.04%) 995 (3.02%) -
dport+tos - - - 72 (0.17%) 76 (0.18%) -
- - - 23 (0.07%) 30 (0.09%) -
- - - 10 (0.03%) 16 (0.05%) -
daddr+dport+fl+per-packet - - - - - -
- - - - - -
5 (0.03%) 1 (0.01%) - - - -
daddr+fl+per-packet - - - - - -
- - - - - -
- - 30 (0.16%) - - -
daddr+dport+per-packet+tos - - - - - -
- - - - - -
- - - 40 (0.12%) 36 (0.11%) -
daddr+dport+per-packet - - - - - -
- - - - - -
9 (0.05%) 9 (0.07%) - 10 (0.03%) 12 (0.04%) -
chksum+daddr+fl+per-packet+tc - - - - - -
- - - - - -
- - 39 (0.20%) - - -
chksum+daddr+fl+per-packet - - - - - -
- - - - - -
- - 2 (0.01%) - - -
daddr+per-packet+tc - - - - - -
- - - - - -
- - 9 (0.05%) - - -
chksum+daddr+per-packet+tc - - - - - -
- - - - - -
- - 9 (0.05%) - - -
daddr+dport+fl+per-packet+tc - - - - - -
- - - - - -
13 (0.08%) 10 (0.07%) - - - -
daddr+per-packet - - - - - -
- - - - - -
5 (0.03%) 1 (0.01%) 185 (0.96%) 5 (0.02%) 15 (0.05%) 47 (0.12%)
dport+per-packet+tos - - - - - -
- - - - - -
- - - - 1 (0.00%) -
chksum+daddr+per-packet - - - - - -
- - - - - -
- - 4 (0.02%) - - 4 (0.01%)
daddr+fl+per-packet+tc - - - - - -
- - - - - -
1 (0.01%) - 6 (0.03%) - - -
chksum+daddr+per-packet+tos - - - - - -
- - - - - -
- - - - - 43 (0.11%)
daddr+per-packet+tos - - - - - -
- - - - - -
- - - 1 (0.00%) 2 (0.01%) 8 (0.02%)
None 973 (4.52%) 986 (5.89%) 1991 (8.89%) 3831 (9.14%) 4446 (10.30%) 4983 (10.88%)
388 (2.24%) 448 (3.30%) 652 (3.35%) 1568 (4.78%) 1803 (5.35%) 1594 (3.92%)
246 (1.44%) 251 (1.88%) 434 (2.25%) 809 (2.52%) 944 (2.87%) 864 (2.14%)