We discussed the usage of system tags for performing vantage point selection of dual-stacked probes. Our exploration reveals how with around 2K dual-stacked probes, RIPE Atlas provides the richest source of vantage points for IPv6 measurement studies. User tags on the other hand are based on a manual process which is largely dependent on proactive participation of probe hosts. We show that user tags tend to become stale over time. This work was presented at the MAT working group at RIPE 72 meeting in Copenhagen.
We helped organize a Dagstuhl seminar on Global Measurements: Practice and Experience from January 04–07, 2016. This was a followup of the seminar on Global Measurement Framework. The second seminar aimed at discussing the practical experience gained with building global measurement platforms. It brought together people who are actively involved in the design and maintenance of global measurement systems, who do research on the data delivered by global measurement systems, and who use data derived from global measurement systems in order to manage networks or services or as input for regulatory decisions.
Our research group is taking an active leadership role in the organization of some key scientific events this year: We are involved in the organization of the demo track of the IFIP/IEEE International Symposium on Integrated Management (IM 2015), which takes place in Ottawa (Canada) in May 2015. We co-chair the technical program committee of the 11th International Conference on Network and Service Management (CNSM 2015), which takes place in Barcelona (Spain) in October 2015.
In this talk, we present a set of tools that we find useful for measurement research and would like to share them with the larger RIPE community. Given the nature of the talk, we will make a live demo of running code snippets using the IPython notebook. A recording of our presentation can be found on Vimeo.
We have demonstrated an implementation of the LOWPAN-MIB at the Low-power Lossy Networks (LLN) plugfest, which took place on Sunday before the 90th IETF meeting. We showed how to access the counters via our Contiki SNMP implementation and Contiki's CoAP implementation. The plugfest was overall very interesting and very well organized. People involved in several interesting projects attended the plugfest and it was nice to get into personal contact. Further details can be found on the plugfest slides.
We gave a talk in which we share our experiences and lessons learned from using the RIPE Atlas platform for conducting measurement research. We describe how subtle rate limits can affect experiment design. We show how calibration of probes from hardware revision down to the firmware version is useful when analyzing measurement results. We describe the usefulness by showing how different hardware revisions affect measurement results. We show how per-hop aggregation mistakes during data analysis can have impacts on measurement results.
We participated in a Dagstuhl seminar on Global Measurements Framework from November 17-20, 2013. The seminar discussed challenges in running large-scale measurement platforms, the associated privacy issues, with a particular focus on how to make it more useful by regulators, Internet service providers, and network users.
Here is another update of our IPv6 traffic. Averaged over the last year, the IPv6 traffic is about 5% of the overall Internet traffic we have. This is pretty good given the fact that only some parts of Jacobs' network currently advertises IPv6. In particular, we have no IPv6 in the student dorms (and the graph clearly suggests that students are a major contributor to our Internet traffic - in particular during finals periods).
Vaibhav Bajpai gave a talk about his work on measuring the effectiveness of happy eyeballs at the RIPE 66 meeting. The IETF has developed solutions that promote a healthy IPv4 and IPv6 co-existence. The happy eyeballs algorithm for instance, provides recommendations to application developers to help prevent bad user experience in situations where IPv6 connectivity is broken. We study the effectiveness of the happy eyeballs algorithm. A recording of the talk can be found on Vimeo.
A RPL-MIB implementation, based on draft-sehgal-roll-rpl-mib-06, has been completed for the Contiki SNMP agent. An AVR Raven mote has been deployed for testing the MIB and is reachable via the following information: Hostname: rpl-mib.eecs.jacobs-university.de Port: 1610 If using the Net-SNMP tools, you can retrieve all the MIB objects by using the following: $ snmpwalk -v 1 -c public udp6:rpl-mib.eecs.jacobs-university.de:1610 1 The current RPL-MIB implementation is hosted under the Jacobs University private enterprise OID number.