Such topic (namely modeling DNS cache behavior) sounds interesting and I’d like
to have a try. By the way, is this topic cared about by this work group?
Guangqing Deng
From: Livingood, Jason
Date: 2013-01-19 00:07
To: Emiliano Casalicchio; John Levine; [email protected]
Subject: Re: [DNSOP] Models of DNS cache behavior?
Oh, and some interesting questions may include a study of:
- How are TTLs changing over time now that so much content is moving to
CDNs
- What are the recursive resolver effects due to more hyper-giants and CDNs
- Study of large scale auth failures in CDN/content providers and the
resulting effects on recursive resolvers (system shocks where QPS load may
go up 100% - 200% in a few minutes)
Jason
On 1/18/13 11:03 AM, "Livingood, Jason"
<[email protected]> wrote:
>If there is anyone out there interested in doing research into DNS
>recursive resolvers and are seeking either (1) access to data (stripped of
>PII of course) and/or (2) modest funding support, please let me know
>off-list.
>
>- Jason
>
>
>
>On 2/20/12 2:27 PM, "Emiliano Casalicchio"
><[email protected]> wrote:
>
>>John,
>>may be the model described in this paper can be useful for you.
>>
>> Yakup Koç, Almerima Jamakovic, Bart Gijsen, "A Global Reference Model of
>>the DNS"
>>
>>presented at DNS-EASY 2011.
>>
>>Here the proceedings:
>>http://www.gcsec.org/sites/default/files/files/dnseasy2011_final.pdf
>>
>>Best
>>Emiliano
>>
>>
>>
>>
>>On 18 Feb 2012, at 02:57, John Levine wrote:
>>
>>> Are there any models of DNS cache behavior, either analytic or
>>> simulations? What I have in mind is something that would help me see
>>> whether I should partition a cache among various kinds of traffic, or
>>> perhaps limit max TTLs, or experiment with replacement strategies.
>>>
>>> For that matter, what's the state of DNS modelling in general?
>>>
>>> I found a paper by Jung et al from 2003 on cache models which starts
>>> by asserting that caches are so big that entries only drop out due to
>>> TTL expiry, did a lot of analysis and simulation, and concluded that
>>> 15 minute TTLs got nearly the same cache benefit of 24 hr TTL.
>>>
>>> A 2010 paper by Alexiou et al. models the Kaminsky DNS poisoning
>>> attack and the port randomizing fix, which is interesting but not what
>>> I'm looking for. (They conclude that the attack is real, and the fix
>>> works OK.)
>>>
>>> Anything else I should be looking at?
>>>
>>>
>>> --
>>> Regards,
>>> John Levine, [email protected], Primary Perpetrator of "The Internet for
>>>Dummies",
>>> Please consider the environment before reading this e-mail.
>>>http://jl.ly
>>> _______________________________________________
>>> DNSOP mailing list
>>> [email protected]
>>> https://www.ietf.org/mailman/listinfo/dnsop
>>>
>>
>>_______________________________________________
>>DNSOP mailing list
>>[email protected]
>>https://www.ietf.org/mailman/listinfo/dnsop
>
>_______________________________________________
>DNSOP mailing list
>[email protected]
>https://www.ietf.org/mailman/listinfo/dnsop
_______________________________________________
DNSOP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dnsop
_______________________________________________
DNSOP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dnsop
