On 16/04/2011, at 5:12 AM, Stephen Kent wrote:

> Robert,
> 
> First let me apologize as I accidentally put in "bytes" where I meant "bits" 
> in the examples I provided. Whoops!  RSA 2K keys are 2K bits, i.e., 256 bytes 
> (not 2K bytes). And, 2K RSA sig is the default, as this matches the key sizes 
> we have already agreed upon for the RPKI certs. Still, a 20-hop path with RSA 
> 2K (bit) sigs exceeds the 4K (byte) max UPDATE size, without any other 
> overhead.  Sorry for the confusion.
> 
>> 
>> - What is the maximum key length possible and how big would be the RSA 
>> signature with such key length ?
> 
> There is no max key size, but it does not make sense to push for bigger RSA
> key sizes, instead of moving to more efficient (in space and computation) sig 
> algorithms. The sig size for RSA is the same size as the key.
> 
> The likely sucessor algs are DSA or EC-DSA, which use smaller keys, but offer 
> secruity equivalent to the larger RSA key sizes. The key sizes for those 
> algorithms yield 128 or 256-bit sigs, under current hash algs.
> 
>> - What are the other path security data and what their size might be min-max.
> 
> I defer to Matt Lepinski for the details of the other data, as he is the 
> author of the BGPSEC protocol doc.

I was doing a similar mental sum in my head and 256 bytes per sig is not a lot. 
As far as I am aware the only other added per-AS part of the attribute is the 
SKI of the public key. Presumably this is also 256 bytes.

SO thats 500 bytes per AS, and for a 20 AS path thats 10K in additional data.

Is my back of the envelope anywhere near in the right ball park?

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

Reply via email to