[Yeti DNS Discuss] [Yeti DNS Operators] additional dns-traffic
ljsong at biigroup.cn
Mon Sep 14 04:30:29 UTC 2015
sorry, I found your mail lying in my junk mailbox, the same case for Shane with BII’s mailbox.
I don’t know why. I cc discuss mailing list instead to involve more discussion on this topic.
I encourage you to do what you proposed and make necessary analysis on 1) whether traffic can spread to
other yeti root servers (with certain NS selection algorithm) , 2) observing if there is any problem for the
query/response, which can help us to improve our testbed
I proposed similar idea to reply the DNS traffic from production network to simulate the real
query load, because we do not have enough resolvers right now. This will be helpful particular
in the beginning of Yeti testbed to enhance its stability and readiness for experiments or production trial.
In the f2f yeti meeting last week, we discuss three way to increase the traffic on yeti,
1) resolve from Universities and research region (Like BUPT as a case)
2) replay some real traffics (as you proposed)
3) use some existing experimental tool like Atlas to generate traffic for special experiment;
> 在 2015年9月8日，23:52，Dmitry Y. Kovalenko <d.kovalenko at msk-ix.ru> 写道：
> Hi all
> We have put some additional dns traffic to yeti lab. You can see it on yeti graph for yeti-ns.ix.ru server
> As I proposed in Prague the scheme is:
> We tap ipv6 dns-queries from our open resolver in Moscow to our test yeti resolver server.
> This server uses yeti root environment, does recursion and drops dns-replies to avoid sending duplicates to dns-clients.
> At the moment our yeti test resolver uses only yeti-ns.ix.ru as a root server, but we can spread traffic to other yeti root servers if it is needed
> Also I can try to add more traffic by using ipv4 dns-queries and do recursion in ipv6 environment where yeti root servers are. Maybe it would work.
> What do you think?
> Kovalenko Dmitrii
> operators mailing list
> operators at lists.yeti-dns.org
ljsong at biigroup.cn
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the discuss