Robin,
What do you think of the idea of making that query server function available to nearby devices?
I ponder this in the NERD draft if one presumes that one loads "near" databases and uses ALT or CONS for "far" networks. A simpler approach, however, given the lower volume of data, would be to receive ALL databases, and keep them in a secondary cache or storage, and then keep the query on the box.
I think this would be an enhancement to the NERD architecture - to allow for this, enabling operators to do it when it made sense to them.
I remain averse to delay. Even the above scheme would require architectural changes, and would impose delay for cache misses, but those cache misses would be extremely small. I also remain averse to mechanisms where an inability to complete a query due to a device failure would cause either yet more delay or a mapping failure. With NERD, the worst that happens is that you use the old mapping for a time.
And indeed I am engineering a solution that does not require host granularity. I believe mechanisms such as HIP and MIPv6 need to be improved to scale to address that problem.
But that's me. Eliot -- to unsubscribe send a message to rrg-request@psg.com with the word 'unsubscribe' in a single line as the message text body. archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg