I just want to point out that I was on a different train.  Discussing the
supposition that query&cahce arch's behave differently that batch pull
archs in either startup or steady state performance.  I was not commenting
on the whole trust/business model discussion.


I was replying to Eric's suggestion that DNS-like systems are an example
of the on-demand architectural choice.

On 12/6/12 10:03 PM, "Eric Osterweil" <[email protected]> wrote:
<snip>
>
>I was hoping we could all see from my quoted text above that this latest
>discussion is about the _architectural_ difference between the on-demand
>soft-state DNS system, and the prefetching replicated state machine of
>RPKI.  These two are fundamentally very different architectural models.
>Your comments about boot states are interesting, but somewhat off topic
>to this post, imho.

Those two choices seem moot from a back of the envelope performance /
behavior perspective (although I would like to run some tests on a 500K
query of a signed reverse DNS ... I don't expect my estimates to get
better) ... With either you had better engineer well with redundant
"gatherers/resolvers", that don't fate share infrastructure, etc ...
Otherwise the hour glass is going to spin for a long time when a DFZ
router tries to validate its RIBs.

Now I return you to your trust/business model discussion ...
Dougm



On 12/7/12 12:34 AM, "Christopher Morrow" <[email protected]> wrote:

>On Thu, Dec 6, 2012 at 11:20 PM, Montgomery, Douglas <[email protected]>
>wrote:
>> I was talking about when the systems designed to support the
>>distribution
>> of authorization information (e.g., RP/RPKI or some DNS based system?)
>> were in steady state ... I.e., they have booted up and done their
>>initial
>> data loads.
>
>somewhere along the train my comment about dns got moved to 'just like
>dnssec' (which I was intentionally not referring to, maybe I should
>have used email as an example instead).  Then in this part of the
>messages I think my DNS comment got moved into 'if you distribute the
>rpki data via dns instead of up/down/left/right/a-b/a-b... the
>repository -> cache protocol'.  I didn't mean it in that sense either
>:(
>
>I think somewhere 5-8 messages back Arturo's note that:
>  1) hosted model is just a crutch
>  2) hosted model isn't intended for everyone to use
>  3) most large ISP or large operations groups are expected to run their
>own CA
>
>coupled with eric's notes that:
>  1) hosted seems fragile for lots of operations
>  2) people should think long and hard about using the hosted model of
>controlling their own fate
>
>gets the general gist of my point: "If you use the hosted model you
>are equivalently outsourcing your Mail/SMTP infrastructure to another
>person, be sure you want to do that..."
>
>apologies for the confusing example :(
>-chris

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

Reply via email to