Re: [DNA] MUST Store Lifetime and Last Refresh time?(draft-pentland-dna-protocol-00.txt)

[Erik Nordmark <erik.nordmark@sun.com>]

Greg,

Thanks for your comments on the draft.

Greg Daley wrote:
> In the draft it states:
> 
>   "The list will be
>    referred to in this document as "DNAPrefixList".  For each prefix the
>    router MUST store the following information:
> 
>    1.  Prefix
> 
>    2.  Prefix length
> 
>    3.  Valid lifetime
> 
>    4.  Time last refreshed"
> 
> In my (existing) implementation I'm storing the expiry time
> of the prefix (which is calculated based on the valid lifetime
> at the RA reception time).

Is this for #3 above? Are you also storing #4?

> Is this insufficient?

Storing the expiry time for the valid lifetime is what's needed for #3.

> I can understand having an indication of when data expires,
> but it seems overkill explicitly pointing out which data types
> have to be internally stored in the host unless there's
> further reasons behind it.
> 
> Would it be possible to make this descriptive and informational
> (being internal to the host) rather than prescriptive,
> using standards language?

The intent is to provide enough *conceptual* data structure descriptions 
so that the protocol can be described in detail, but not constrain how 
an implementation actual does things as long as the externally visible 
behavior of the implementation is consistent with the specification.

There is text in section 5 in RFC 2461 which says this, so perhaps we 
can just borrow that text.

    Erik