Re: [lisp] Fwd: [rrg] [GROW] Operational experience with cache basedmapping ID

Dino,

|Tony, you know how hierarchy works. You follow coarse prefixes as you  
|go up and more specifics as you go down. See draft-meyer-lisp- 
|cons-04.txt for details. And refer to the VA stuff Paul is working on.

I do indeed know how hierarchy works.  I know that a caching element in a
hierarchy that overruns its working set is going to thrash.  I know that you
can only scale if you distribute the work (in this case caching) down the
hierarchy towards the leaves.  I know that redirecting those requests
upwards is simply going to cause the collapse of higher layers. 

I also know that it's trivial to attack caches through a random walk of the
namespace.  I also know that it's trivial to trigger that remotely if there
is any way to get a response to a random incoming source.  

I know that any proposal needs to address these issues, not duck them.

|> Second, there have been numerous alternatives discussed: full map,  
|> DHT, and
|> hybrids.
|
|Right, and I think many of these are misunderstood because they are  
|not brought to the level of detail to find the bugs.

I think that many people are not willing to have the architectural
discussions necessary because they find it more fun to jump into the
architecturally irrelevant details.

Tony

_______________________________________________
lisp mailing list
lisp@ietf.org
https://www.ietf.org/mailman/listinfo/lisp