Internet2 NetFlow: Weekly Reports: Week of 20091130

  1. Introduction
  2. Bulk TCP
  3. Full Data Set

Introduction

You are looking at the weekly Abilene network usage report for the week of 20091130 produced from NetFlow records. The view of the whole network as a single traffic-relaying unit is presented. More formally, data from all interior circuits (those connecting two Abilene routers) were discarded while all the rest of the data were merged to create this view.

During this week, there were no missing data days.

The data are split into two sections: bulk TCP data and the full data set. A "bulk TCP" flow is defined as a TCP flow that transferred more than 10MB of data. The first section only concerns these data. The second section studies the overall traffic composition.

All the numbers in this report are hyperlinked to plots that show their history (e.g., clicking on the percentage of octets of NNTP traffic will bring up a time-series plot that shows the history of this parameter).

Bulk TCP

During this week, bulk TCP traffic comprised 26.23% of octets and 0.88% of packets of the full data set traffic.

The distribution of bulk TCP throughputs is the most important piece of data in this report. Cumulative distribution function plots (1-CDF vs. throughput in bits/second) in semi-log and log-log scales are as follows:
[Bulk TCP throughputs (semi-log scale).] [Bulk TCP throughputs (log-log scale).]

Distribution of the amount of data transferred (in semi-log and log-log scale, 1-CDF vs. total trasfer size in octets) is presented below. It should be recognized that NetFlow collection mechanism is always configured so that flows (in the accounting sense) cannot last longer than a certain period of time. Therefore, the distribution of transfer sizes is to a certain extent skewed in the upper part.
[Bulk TCP transfer sizes (semi-log scale)] [Bulk TCP transfer sizes (log-log scale).]

The distribution of durations of bulk TCP flows (in seconds) is as follows (you may notice the cut-off phenomenon mentioned above):

[Bulk TCP durations distribution.]

The following table shows actual values from the above distribution plots that correspond to characteristic values (such as median, 90%, max, etc.).

Table 1. Selected Points from Distribution Graphs (Bulk TCPs)

Percentile Throughput (b/s) Durations (s) Size (octets)
1 1.397M 2 10.08M
5 1.491M 8 10.41M
10 1.600M 15 10.93M
50 3.137M 57 17.27M
90 14.71M 59 49.55M
95 25.99M 59 74.66M
99 75.66M 59 177.9M
99.9 183.9M 59 535.3M
99.99 446.5M 59 1.762G
99.999 899.3M 113 3.552G
100 36.58T 4121998 423.5G

We compute average packet size of each flow by dividing the number of octets in a flow by the number of packets. Distribution of average sizes of packets belonging to bulk TCP flows is as follows:

Table 2. Packet Sizes (Bulk TCP)

Packet Size Packets
Small (<100B)91.42% 4.091T
Medium (100-1400B)0.83% 37.06G
Large (1401-1500B)7.75% 346.7G
Jumbo (>1500B)0.00% 44.50M
Total100.00% 4.475T

We show what applications transfer large amounts of data in the following table. Note that this is bulk TCP traffic only; full data set usage is presented in the next section.

Table 3. Aggregated Application Types (Bulk TCP)

Traffic Type OctetsPacketsFlows
Data Transfers34.94% 194.6T 51.31% 2.296T 38.64% 7.618M
Encrypted Traffic9.14% 50.94T 0.87% 38.97G 5.58% 1.099M
Advanced Apps3.30% 18.39T 0.29% 12.78G 4.20% 828.8k
File Sharing2.99% 16.65T 5.20% 232.7G 2.32% 456.5k
Measurement1.04% 5.799T 0.11% 4.940G 0.20% 38.77k
Misc0.76% 4.231T 0.07% 3.071G 1.09% 215.5k
Games0.21% 1.152T 0.02% 823.1M 0.24% 47.40k
Audio/Video0.10% 568.9G 0.01% 400.8M 0.20% 38.65k
Unidentified47.52% 264.7T 42.12% 1.885T 47.54% 9.373M
Total100.00% 557.1T 100.00% 4.475T 100.00% 19.71M

The following are the fastest 10 measurement flows with unique source and destination AS numbers (i.e., for any given pair of source and destination AS numbers, no more than one fastest flow is shown).

Table 4. Fastest Bulk TCP Measurement Flows with Unique AS Source and Destination

Throughput (b/s)Packet size (bytes)Duration (s)Src ASDest ASApplication type
976.5M150020UIUC [38]Boston U [111]Iperf
975.1M150010Unknown [32361]U Chicago [160]Iperf
970.8M147320Unknown [32361]VANDERBILT [7212]Iperf
965.4M150020Unknown [32361]SDSC [195]Iperf
962.0M146412INDIANAGIGAPOP [19782]Unknown [32361]Iperf
953.9M150020Unknown [32361]Boston U [111]Iperf
925.1M150020Unknown [32361]U Wisconsin [59]Iperf
922.9M150020UIUC [38]U Chicago [160]Iperf
914.5M150020UNL [7896]Purdue [17]Iperf
904.7M146414Boston U [111]Unknown [32361]Iperf

The following are the fastest 10 non-measurement flows with unique source and destination AS numbers (i.e., for any given pair of source and destination AS numbers, no more than one fastest flow is shown). When unable to determine the application type, we give the source and destination port numbers.

Table 5. Fastest Bulk TCP Non-measurement Flows with Unique AS Source and Destination

Throughput (b/s)Packet size (bytes)Duration (s)Src ASDest ASApplication type
153.0G112Unknown [0]Merit [237]HTTP
1.008G150014Unknown [32361]SDSC [195]5016 -> 5016
983.9M147213Unknown [32361]VANDERBILT [7212]5012 -> 5012
938.5M146459Abilene [11537]BWI-GIGA-POP [10886]Shoutcast
795.8M150014Fermi National Accelerator Lab [3152]UNL [7896]38537 -> 58934
756.5M150012Unknown [32361]Stephen F. Austin State U [3634]5018 -> 5018
675.0M150016U Wisconsin [59]UNL [7896]46326 -> 37430
623.1M150028Unknown [32361]Universiy of California, San Diego CA [7377]47772 -> 40736
601.2M146420U Florida [6356]Stephen F. Austin State U [3634]5015 -> 5015
585.0M150028U Wisconsin [59]Universiy of California, San Diego CA [7377]38799 -> 40760

We also compute the average concurrency of bulk TCP flows for the week (by adding durations of all captured flows and dividing the result by the by the duration of the week). This week's average number of concurrent bulk TCP flows: 1.593k.

Full Data Set

In addition to bulk TCP flows data, we provide statistics that characterize the overall composition of the complete data set (everything that transited the Abilene network this week).

The following table describes what kinds of traffic went through the network (multiple applications are aggregated into classes):

Table 6. Aggregated Application Types (Full Data Set)

Type OctetsPackets
Data Transfers37.23% 790.7T 1.56% 7.917T
Encrypted Traffic5.09% 108.1T 0.45% 2.298T
Audio/Video2.14% 45.39T 7.59% 38.62T
Advanced Apps1.36% 28.88T 0.01% 30.92G
File Sharing1.23% 26.15T 0.08% 402.0G
Misc1.16% 24.57T 0.11% 561.8G
Measurement0.46% 9.855T 0.34% 1.735T
Games0.31% 6.506T 0.26% 1.319T
Unidentified51.03% 1.083P 89.61% 456.0T
Total100.00% 2.124P 100.00% 508.9T

This table is available additionally in the following more verbose version (no applications are aggregated into classes, but class composition is shown):

Table 7. Detailed Application Types (Full Data Set)

Traffic type OctetsPackets
Data Transfers
HTTP
Rsync
FTP
NNTP
---
35.85%
0.90%
0.34%
0.14%
---
761.4T
19.07T
7.235T
2.985T
---
1.55%
0.00%
0.00%
0.00%
---
7.883T
18.84G
8.129G
6.365G
Encrypted Traffic
SSH
HTTPS
IPsec ESP
IPsec AH
IPsec IKE
---
2.34%
2.10%
0.59%
0.06%
0.00%
---
49.75T
44.57T
12.63T
1.193T
11.13G
---
0.01%
0.02%
0.14%
0.28%
0.00%
---
55.63G
84.91G
710.3G
1.447T
51.63M
Audio/Video
Any-Source Multicast
Real Player
Single-Source Multicast
Windows Media
Backbone Radio
H.323 Signaling
StreamWorks
Camarades webcams
Subset of VoIP
---
1.83%
0.17%
0.11%
0.01%
0.00%
0.00%
0.00%
0.00%
0.00%
---
38.92T
3.600T
2.308T
317.9G
88.12G
87.03G
39.65G
20.04G
4.988G
---
7.13%
0.05%
0.41%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
---
36.28T
243.4G
2.091T
328.1M
114.9M
116.4M
52.64M
32.63M
10.60M
Advanced Apps
UNIDATA LDM
McIDAS
BBCP
GsiFTP
BBFTP
IBP
---
1.21%
0.09%
0.05%
0.00%
0.00%
0.00%
---
25.76T
1.885T
1.168T
47.30G
15.08G
5.444G
---
0.01%
0.00%
0.00%
0.00%
0.00%
0.00%
---
27.73G
1.521G
1.487G
105.8M
63.41M
9.849M
File Sharing
Audiogalaxy
Hotline
Shoutcast
BitTorrent
eDonkey2000
FastTrack
Gnutella
WinMX
Carracho
Freenet
Blubster
Neo-Modus
Direct Connect++
---
0.70%
0.27%
0.13%
0.08%
0.05%
0.01%
0.01%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
---
14.78T
5.690T
2.729T
1.672T
973.1G
120.3G
109.5G
51.32G
11.41G
5.095G
3.059G
303.6M
264.5M
---
0.05%
0.00%
0.00%
0.00%
0.03%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
---
236.1G
4.981G
4.447G
2.013G
153.9G
177.2M
240.3M
76.65M
20.09M
6.560M
39.84M
887.8k
421.2k
Misc
Mail
DNS
Squid
X11
Port 0
MS Windows
NTP
AFS
IRC
RTIP
NFS
SOCKS
IDENT
Telnet
SNMP
AOL AIM
RPC Portmapper
---
0.84%
0.10%
0.08%
0.04%
0.03%
0.03%
0.01%
0.01%
0.01%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
---
17.78T
2.171T
1.633T
797.1G
734.3G
566.4G
250.2G
248.4G
131.2G
68.31G
67.12G
31.55G
31.35G
22.05G
20.95G
17.16G
505.8M
---
0.06%
0.00%
0.00%
0.00%
0.00%
0.04%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
---
314.1G
18.04G
2.562G
1.245G
1.557G
218.3G
3.287G
747.4M
473.0M
702.8M
132.4M
60.78M
78.41M
181.3M
147.8M
26.23M
3.279M
Measurement
Iperf
ICMP
IPMP
---
0.35%
0.08%
0.04%
---
7.412T
1.670T
772.3G
---
0.00%
0.19%
0.15%
---
8.716G
957.0G
769.5G
Games
DirectX
Battlenet
Quake
Half-Life
Starsiege Tribes
Asheron
Spy Arcade
---
0.17%
0.07%
0.03%
0.02%
0.00%
0.00%
0.00%
---
3.587T
1.580T
630.4G
501.4G
94.35G
68.28G
44.52G
---
0.01%
0.18%
0.07%
0.00%
0.00%
0.00%
0.00%
---
68.82G
903.5G
342.9G
3.539G
389.5M
128.0M
108.2M
Unidentified
Unidentified
---
51.03%
---
1.083P
---
89.61%
---
456.0T
Total
Total
---
100.00%
---
2.124P
---
100.00%
---
508.9T

The following table summarizes use of most popular IPv4 protocols:

Table 8. IP Protocols Distribution (Full Data set)

Protocols OctetsPackets
ICMP[1]0.08% 1.670T 0.19% 957.0G
IGMP[2]0.03% 662.7G 0.14% 710.2G
IP-ENCAP[4]0.04% 745.2G 0.10% 500.7G
TCP[6]70.65% 1.500P 2.16% 10.99T
UDP[17]5.07% 107.7T 0.84% 4.283T
IPv6[41]0.12% 2.472T 0.58% 2.937T
GRE[47]0.47% 10.05T 0.42% 2.131T
ESP[50]0.59% 12.63T 0.14% 710.3G
AX.25[93]0.08% 1.693T 0.28% 1.427T
PIM[103]0.07% 1.451T 0.32% 1.610T
IPMP[169]0.04% 772.3G 0.15% 769.5G
Other22.77% 483.5T 94.69% 481.9T
Total100.00% 2.124P 100.00% 508.9T

We compute average packet size of each flow by dividing the number of octets in a flow by the number of packets. Distribution of (average) packet sizes is as follows:

Table 9. Packet Sizes (Full Data Set)

Packet Size Packets
Small (<100B)99.74% 507.6T
Medium (100-1400B)0.09% 448.9G
Large (1401-1500B)0.17% 856.3G
Jumbo (>1500B)0.00% 3.337G
Total100.00% 508.9T

We only track DSCP values for which special treatment was defined by Internet2 QoS working group (and the default of DSCP=0):

Table 10. Important DSCP Values (Full Data Set)

Type OctetsPackets
Best effort [DSCP=0]74.34% 1.579P 3.06% 15.59T
Scavenger [DSCP=8]0.36% 7.649T 1.19% 6.046T
EF [DSCP=46]0.27% 5.797T 1.05% 5.334T
Other25.02% 531.5T 94.70% 481.9T
Total100.00% 2.124P 100.00% 508.9T

We collect statistics about ECN-capable traffic:

Table 11. ECN-Capable Traffic

Type OctetsPackets
ECN-Capable19.10% 405.6T 79.06% 402.3T

To facilitate detection of emerging applications, we present statistics about frequently encountered unidentified port numbers (no distinction is made in this table between TCP and UDP):

Table 12. Frequent Unidentified Ports

Port OctetsPackets
19356.02% 127.8T 0.03% 159.8G
600110.96% 20.35T 0.00% 22.29G
164020.67% 14.30T 0.00% 17.28G
150000.53% 11.31T 0.00% 15.46G
330010.35% 7.375T 0.00% 5.053G