On 28 Apr 2014, at 23:07, Douglas Otis wrote:

> 
> 
> On Apr 28, 2014, at 2:32 PM, Trevor Freeman <[email protected]> 
> wrote:
> 
>> I spoke to soon. While the US government domains  is signed, the actual web 
>> site is not in many cases.
>> For example:
>> www.dhs.gov is a cname entry www.dhs.gov.edgekey.net which is unsigned.
>> This is in turn a CNAME to another unsigned domain
>>  
>> www.dhs.gov.edgekey.net is a CNAME to e6485.dscb.akamaiedge.net
> 
> Dear Trevor,
> 
> Yes, there are many such CNAME wildcards in use.  A poor practice from many 
> perspectives.  
> 
> Comcast has been helpful in vetting DNSSEC use. See http://dns.comcast.net.  
> http://tools.ietf.org/html/draft-start-tls-over-dns-00 solves many deployment 
> and privacy issues.  DNSSEC should be considered a much needed work in 
> progress.
> 
> Regards,
> Douglas Otis

PKI already suffers from the fact that its trust anchors do not align well (or, 
usually, at all) with the trust factors that underpin applications/services. 
Cloud architectures seem to me to exacerbate that state of affairs, because 
they often introduce a further level of abstraction in the certificate 
structure. I've seen this in two forms, each of which bears similarities to the 
akamai example Trevor cited above:

1 - you visit "nakedfurries.com" (sorry, Warren ;^)  ) but because their 
website is hosted, you see certificates for "cloudstorm.net" instead or as 
well, which doesn't contribute anything meaningful to the user's assessment of 
trust;

2 - the cloud service provider adds seemingly random qualifiers to the 
certificate label ( e.g. fjdo7pspddlfhe544gr6g.servicename.com). Again, this 
doesn't add to the user's trust assessment, and can look suspiciously as though 
the cert label is being used as the vehicle for a unique tracking ID.

Yrs.,
Robin

> 
>  
>> From: perpass [mailto:[email protected]] On Behalf Of Trevor Freeman
>> Sent: Monday, April 28, 2014 2:17 PM
>> To: Noel David Torres Taño; [email protected]
>> Subject: Re: [perpass] Is DNSDEC a viable technology for perpass?
>>  
>> Hi Noel,
>>  
>> If DNNSEC is used in corporations, that may be an interesting data point but 
>> perpass is specify looking at the interne so it does not help much.
>>  
>> I understand they could be some benefit to adding some other filter to the 
>> data but the number to try and try to add a better quality metric. But 
>> absent that, the number is what is it. Happy to have the discussion on how 
>> we would consider what to filter on and maybe Verisign could provide more 
>> attributes with the data for use to mine the information.  
>>  
>> I did some ad-hoc research and amongst the prominent internet services or 
>> financial institutions, the seems little evidence of DNSSEC.  The only 
>> bright spot seemed to be government web sites, though here the deployment 
>> was still inconsistent in that government agencies have many web sites not 
>> part of the base domain and these were often not signed.
>>  
>> Trevor
>>  
>> -----Original Message-----
>> From: perpass [mailto:[email protected]] On Behalf Of Noel David 
>> Torres Taño
>> Sent: Monday, April 28, 2014 1:02 PM
>> To: [email protected]
>> Subject: Re: [perpass] Is DNSDEC a viable technology for perpass?
>>  
>> El lun, 28-04-2014 a las 18:38 +0000, Trevor Freeman escribió:
>> > We have a range of technologies in the toolkit to address issues
>> > identified by perpass.
>> >
>> > 
>> >
>> > One of the candidate technologies is DNSSEC. At a technology level it
>> > has much to commend it.
>> >
>> > 
>> >
>> > The vast majority of critical TLDs are signed, so another good point
>> > in its favor.
>> >
>> > 
>> >
>> > However when you look at the next tier down, the statistics point to a
>> > problem.
>> >
>> > 
>> >
>> > According to the Verisign labs scoreboard, 340K+ domains in the .com
>> > namespace are secured by DNSSEC
>> >
>> > http://scoreboard.verisignlabs.com/
>> >
>> > 
>> >
>> > If you express that number as % that is about 0.4% and the growth
>> > trend is about 0.1% per year
>> >
>> > http://scoreboard.verisignlabs.com/percent-trace.png
>> >
>> > 
>> >
>> > The trend seems about 2 orders of magnitude below where we need to be
>> > for DNSSEC to be viable in a realistic timescale.
>> >
>> > 
>> >
>> > Am I misinterpreting the data? If not, then do we have consensus on
>> > what is blocking deployment?
>> >
>> > 
>> >
>> > Trevor
>> >
>> > 
>> >
>> Which are the numbers for .org ?
>>  
>> This one should have a little percentage of garbage, parked domains, etc. 
>> Moreover, it is kess used by corporations with large IT departments and more 
>> used by small organizations like Libre Software projects.
>>  
>> And it is very important to trust the software you download.
>>  
>> Regards
>>  
>> Noel
>> er Envite
>>  
>> _______________________________________________
>> perpass mailing list
>> [email protected]
>> https://www.ietf.org/mailman/listinfo/perpass
> 
> _______________________________________________
> perpass mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/perpass

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

_______________________________________________
perpass mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/perpass

Reply via email to