Summary

 

The "Local Perspective" Work party held its kickoff meeting on Thursday, 30 April 2020. Participation was good with approximately 20 people. There was interest in volunteering for the role of Work Party Leader, however, it was decided to postpone the selection of a leader until the next meeting, allowing participants to learn more about the topic. The Statement of Work was presented and a few topics were discussed among the group. For future meetings, a doodle poll will be initiated to determine one or two standing times for monthly calls. Plans to hold a work party meeting as part of the ICANN 68 (Kuala Lumpur, virtual) meeting were dismissed as the work party preferred to decide on a time slot on their own versus being tied to local time in KL.

 

 

Action Items

 

1. Staff will conduct a poll to determine a standing meeting time for monthly calls

2. Work Party members are encouraged to volunteer to act as work party leader

3. Work Party members are encouraged to continue discussions on the RSSAC Caucus mail list on new or existing topics related to this work party.

4. Ken Renard and staff will send out meeting summary to RSSAC Caucus mail list

 

 

Topics discussed

 

  1. Limiting scope

Duane brought up the point that the work party might be attempting too much to come to a successful conclusion. While implementations of a full set of tools and a data collection infrastructure would be nice to have, these seem like goals that can be put off for future work. The work party should concentrate on defining metrics to achieve the goals of the intended audience.

  1. Other research

Jaap contributed that CAIDA has been conducting similar research to date and we should reference that work and leverage its results as appropriate. A reference to any publications from this work would be useful for the group]

  1. Purpose of this work

The purpose of this work is broken in to 2 categories: advising the RSS on placement of new instances and giving the end-user a measurement of their perspective of the RSS. Advising the RSS can only be successful if there are a large and "well-distributed" set of measurements.

  1. Measurement of endpoint to RSS versus via configured resolver

The work party had a discussion on the desire for a tool to measure direct perspective of the RSS versus measuring perspective via a configured recursive resolver. The tool(s) may be run on a wide variety of platforms such as mobile devices, desktop hosts, data center servers, enterprise recursive resolvers, etc. Direct measurement to the RSS would be interesting for potential placement of new RSS instances, but location of the measurement (such as mobile endpoints) need to be considered. Measurement via the configured recursive resolver would be useful for anyone to get a perspective of their enterprise/ISP configuration.

 

 

Agenda topics for next meeting

 

  1. Select a work party leader
  2. Enumerate end-user use cases:
    1. Researcher/curious user in a University or Enterprise network
    2. Enterprise IT staff (recursive resolver operators)
    3. Topologically-remote network.
    4. Well-connected data centers (Google/CloudFlare) good for comparison?
    5. Home users (testing ISP versus public DNS perspective of root)
    6. RSOs
  3. Discussion of CAIDA research
  4. What set of measurements would be useful for testing perspective directly to RSS?  Let's get an initial list to ground general thoughts on the set.  We will revisit this periodically.
    1. Local IP (ASN), geo-location?, RSS RTT (v4/v6/udp/tcp), RTT to first-hop router, RTT to well-known Internet infrastructure, traceroutes to RSS
  5. What set of measurements would be useful for testing perspective of the RSS via the local resolver?  Let's get an initial list to ground general thoughts on the set.  We will revisit this periodically.
    1. IP (ASN) of RR, RTT to RR, RSS RTT (v4/v6/udp/tcp), RTT to first-hop router, RTT to well-known Internet infrastructure, traceroute (from where?)
    2. Compare RTT for cache miss versus cache hit
    3. Compare RTT of configured RR versus well-established resolver (google/cloudflare/etc)

 

Ken Renard

S&TCD Contractor – ICF

Sustaining Base Network Assurance Branch 

C5ISR Center, Space and Terrestrial Communications Directorate

Office:  443-395-7809

kenneth.d.renard.ctr@mail.mil