23C3 - 1.5
23rd Chaos Communication Congress
Who can you trust?
| Speakers | |
|---|---|
|
Elisa Jasinska |
| Schedule | |
|---|---|
| Day | 4 |
| Room | Saal 2 |
| Start time | 17:15 |
| Duration | 01:00 |
| Info | |
| ID | 1644 |
| Event type | Lecture |
| Track | Science |
| Language | English |
| Feedback | |
|---|---|
|
Did you attend this event? Give Feedback |
sFlow
I can feel your traffic
The explosion of internet traffic is leading to higher bandwidths and an increased need for high speed networks. To analyze and optimize such networks an efficient monitoring system is required. The sFlow standard describes a mechanism to capture traffic data in switched or routed networks. It uses a sampling technology to collect statistics from the device and is for this reason applicable to high speed connections (at gigabit speeds or higher).
sFlow is a sampling mechanism suitable for collecting traffic data of high speed networks. A relative small stream of sFlow datagrams provides enough information for statistical analysis of traffic flows.
An Internet Exchange (IX) interconnects various network providers, for example ISP's. The Amsterdam Internet Exchange (AMS-IX) is by its amount of traffic the biggest Internet Exchange in the world. To give the AMS-IX members more insight into their peering traffic and provide information to optimize the network structure, AMS-IX is using sFlow for its traffic analysis.
A throughput average of more then 100 Gb/s gets analyzed by an open source software developed in perl. Due to sFlow providing a whole captured packet (layer 2 - 7) AMS-IX also provides information for example on the growth (or lack off) of IPv6. Information about the sort of traffic might be misunderstood and politically misused therefore AMS-IX restrains itself to layer 2 and the developed software doesn't decode the provided packets above L2.
This topic will contain an introduction to the sFlow sampling mechanism, the information provided by the sFlow datagrams and how they can get analyzed. Besides that, existing tools and the software developed and used at AMS-IX will be presented, and some results of the analysis will be shown.
The software will be hopefully also deployed at the 23C3, and finally we will also see statistics about the network traffic of the conference.