[Yeti DNS Discuss] Note for Yeti Virtual meeting 20150907
ljsong at biigroup.cn
Mon Sep 14 07:05:43 UTC 2015
Last week we hold a online meeting with our Yeti participants(3 coordinators and 5
yeti root operators). The major goal of yeti virtual meeting is to update and share
information to enhance the transparency and communication among Yeti participants.
I made a brief note of that meeting for our information. if I miss anything, please make
necessary adding to this note for those who participated this meeting.
1. Shane and Davey give a presentation introducing current status of Yeti DNS project and
some ideas of future experiments. (please check the slides in previous mail).
2. Right now we do not have enough resolvers and traffic in Yeti testbed. Davey Introduced
a case deploying dual-stack RDNS and Yeti root instance in teaching buildings in an university
in CERNET2 (BUPT CASE). Although it only contributes less than 1 qps into Yeti root system,
but it views as a good example to expand our Testbed. (Soon a Yeti case study in BUPT will be
Note that it is not a quick approach to ask universities or research labs to contribute *considerable*
traffic (say 1k-10k qps). We also discussed other options like mirroring traffic and using atlas
to produce some experimental traffic. we should study case by case somehow.
3. Currently there is only 11 root server in Yeti testbed, not enough authoritative servers than we
expected. It will not fulfill the requirement of some operational experiments.
One idea is straightforward that allow current yeti root operators to run morn than one root NS
server which is proposed by Dahu Group (Stephane Bortzmeyer). From both research and
operational aspects it sounds OK, given that in IANA root system VeriSign runs A and J as well.
The only concerns are : 1) the diversity of system and network; 2) the opportunity for future
participants given 25 is our estimate number of of Yeti root NS server. The conclusion on that
issue is that we are going to draft a *on-boarding document* to make clear the policy of adding
new yeti root operators and servers.
4. Now we have real traffic from our participants. There will be more in the future. In the meeting we
decide the most priority is to keep the testbed stable and consistent, to make it “production-ready”.
We should be more prudent to conduct any changes or experiments before they are fully discussed
and tested. < After virtual meeting , coordinators spend one day working on the system>
5. We also discuss the ways how to enhance the cooperation among Yeti participants.
* Find a way to dynamically share the contact information of participants among us. (other than google doc)
* We need Yeti Wiki, Yeti RT system and making document in collaborative way (other than google doc)
* Calendar for experiments
> 在 2015年9月8日，11:42，Davey(宋林健) <ljsong at biigroup.cn> 写道：
> Hello ,
> The attached is the slides we show in the virtual meeting. Welcome discussions.
> http://yeti-dns.org/resource/2015-09-07-yeti-virtual-meeting.pdf <http://yeti-dns.org/resource/2015-09-07-yeti-virtual-meeting.pdf>
> Davey Song(宋林健)
> BII Lab
> ljsong at biigroup.cn <mailto:ljsong at biigroup.cn>
ljsong at biigroup.cn
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the discuss