Re: [hostmaster-staff] Re: MIR proposal / reservation revisited?
PS To answer Wilfreds question, it is not reservation revisited it is aggregation revisited ;)
Stephen, in my opinion (from a technical persepctive only), there are two ways of attacking the problem: - 1) to clean up the legacy routes, you have to re-number (which I guess can be done within your address pool(s) - 2) to prevent it from happening again, someone, soemwhere, has to reserve a "big" block of addresses which would support a) additional demand by existing customers, and b) aggregation according to your network's internal structure Item 2) is where my notion of "reservation" is coming from. Do I still get it (completely) wrong? Btw, doing only one of those two is probably not going to work... Wilfried.
I have to say purley from a thechnical perspective this has merit.
PS To answer Wilfreds question, it is not reservation revisited it is aggregation revisited ;)
Stephen,
in my opinion (from a technical persepctive only), there are two ways of attacking the problem:
- 1) to clean up the legacy routes, you have to re-number (which I guess can be done within your address pool(s)
Ok this is the legacy IP stuff that has been there for upto 10 years and the mess is partly from the ammalgamation of 17 registries (uunet) Worldcom networks, MCI networks, AOL networks (legacy) and a number of other smaller networks such as Pipex, NLnet and others. Renumbering this would be a nightmare but not impossible, the biggest problem is getting customers to renumber, as i am sure you are well aware. We could look at trying to renumber by customer churn into the new space and hand it back once free. The main reason is not to fix the current mess but to avoid more mess in the future, that said we are trying to reaggregate teh current routing tables to bew more efficient.
- 2) to prevent it from happening again, someone, soemwhere, has to reserve a "big" block of addresses which would support a) additional demand by existing customers, and b) aggregation according to your network's internal structure
Ok so what you are describing is in simplistic terms the MIR concept. The MIR would have a big block of addresses not reserved but allocated to it. This would be based on networks size LIR number and other constraints. Aggregation would be accross the 1 AS (AS702) and new assignments would be made out of the new space. Though the allocation (sub-allocation/partitioning) would be made by the MIR to the LIR the LIR would talk to RIPE for assignment approval. When the LIR reached 80% and only when then more space could be applied for to the MIR and get a block of addresses from the same CIDR preferably applying the NCC method of use last so natural holes are created next to the initial allocation.
Item 2) is where my notion of "reservation" is coming from. Do I still get it (completely) wrong?
No i think you understand the issues just not the solution, reservation is a bad word and bad practice in the current climate.
Btw, doing only one of those two is probably not going to work...
Wilfried.
participants (2)
-
Stephen Burley -
Wilfried Woeber, UniVie/ACOnet