It is a statement and decision of one-year extension for Yeti DNS Project, to the end of 2019.
Yeti DNS Project was started on 8th of May 2015. It was planned as a three-year project with a list of research agenda. However, It spent nearly one year to build the stable DNS root testbed. At that time it was expected the project would be extended a little bit , to the end of 2018. In the past 3 years, several planned Yeti experiments were performed, the system was shown to work, interesting findings were found and Yeti mailing list became a platform to discuss DNS root and IPv6 DNS. (More information was recorded in http://yeti-dns.org/documents.html and I-D: draft-song-yeti-testbed-experience )
It seems that we can safely end Yeti project by the end of 2018 gracefully. But after several round of discussions with Yeti coordinators and sponsors, we think one year extension is needed due to following reasons:
1) There is an important Yeti experiments which need time to be performed: KSK algorithm rollover. The KSK algorithm rollover will be a little challenging, because it is not done in real operational root system. And it may be more complex than KSK rollover due to unknown readiness for various implementations for rolling to ECDSA , especially considering the readiness monitoring and backward compatibility of rollover which IMHO are lessons from ICANN’s KSK rollover.
2) Some Yeti sponsors hope Yeti DNS project be continued with some major changes. So we need time to “Re-charter” Yeti project with clear goal. It was proposed to change the project in a significant way. One is for operational use, another is something more experimental breaking the rule of keeping root zone intact. After resolving the distractions and ambiguity of Yeti in 2017, all coordinators had reached the consensus that the best value of Yeti is being a testbed for research purpose. So we need time to carefully consider what can we do in Yeti DNS testbed next, other than to build a system in a rash.
One candidate input is to try Blockchain technology in Yeti testbed to change how DNS zone organized nowadays. It is inspired by recent discussions in DNSOARC mailing list on “blockchain DNS” and seciton 3.12 in RFC8324. We can introduce a new testing TLD for that purpose without breaking current Yeti root services. It is still open for other proposal candidates of how to change Yeti root significantly.
3) In the meanwhile of Yeti Re-chartering, we need time to look for more sponsors and resource for future Yeti. We need stable funding, full time employers and commitment from our partner.
If both 2# and 3# are successful, then we can proceed more confidently.
Thanks for all sponsors, operators and participants of Yeti project in past years. I hope we can work together in future. Please let us your thoughts and suggestions.
Davey Song on behalf of Yeti coordinators