2006-07 New Policy Proposal (Minimum IPv4 Assignment Window)
PDP Number: 2006-07 Minimum IPv4 Assignment Window Dear Colleagues, A new RIPE Policy has been made and is now available for discussion. This proposal suggests the minimum Assignment Window (AW) available to LIRs should be raised from zero (0) to /21 (2048 IPv4 addresses). Because the sub-allocation policy references the AW policy, the sub-allocation policy also needs to be updated. This proposal suggests that the maximum sub-allocation should be kept at /20 (4096 IPv4 addresses). You can find the full proposal at: http://www.ripe.net/ripe/policies/proposals/2006-07.html We encourage you to review this proposal and send your comments to <address-policy-wg@ripe.net> before 10 October 2006. Regards Filiz Yilmaz RIPE NCC Policy Development Officer
Dear Colleagues, Filiz Yilmaz wrote:
PDP Number: 2006-07 Minimum IPv4 Assignment Window
Dear Colleagues,
A new RIPE Policy has been made and is now available for discussion.
This should read "A new RIPE Policy Proposal has been made and is now available for discussion". Apologies for duplicates and any confusion. Kind regards, Filiz Yilmaz RIPE NCC Policy Development Officer
This proposal suggests the minimum Assignment Window (AW) available to LIRs should be raised from zero (0) to /21 (2048 IPv4 addresses). Because the sub-allocation policy references the AW policy, the sub-allocation policy also needs to be updated. This proposal suggests that the maximum sub-allocation should be kept at /20 (4096 IPv4 addresses).
You can find the full proposal at:
http://www.ripe.net/ripe/policies/proposals/2006-07.html
We encourage you to review this proposal and send your comments to <address-policy-wg@ripe.net> before 10 October 2006.
Regards
Filiz Yilmaz RIPE NCC Policy Development Officer
Hi, On Tue, Sep 12, 2006 at 10:18:04AM +0200, Filiz Yilmaz wrote:
This proposal suggests the minimum Assignment Window (AW) available to LIRs should be raised from zero (0) to /21 (2048 IPv4 addresses).
I'm not fully happy with this proposal, because I think a /21 is too big - that's half the amount of addresses a new LIR receives, and the general assumption "a new LIR doesn't yet know how the game goes" still holds. Argueing "if people don't know what they are doing, they will be in for a hard time when they come for a new allocations" is true, but will hit the wrong people - it means the LIR will have done assignments to end users that will be very difficult to revoke (if done inappropriately), and *new* customers of that LIR might have to wait for a long time for new addresses (because the new allocation is delayed). So the customers get hurt because of sloppiness elsewhere in the process. I'd propose a minimum AW of /24, to be raised to a /22 if the LIR has sent one? two? person(s) to a LIR training course (and the trainer is convinced that these people have listened and understood the rules). (Because having good training doesn't help if people don't use it) So... let's have some discussion. Gert Doering -- NetMaster -- Total number of prefixes smaller than registry allocations: 98999 SpaceNet AG Mail: netmaster@Space.Net Joseph-Dollinger-Bogen 14 Tel : +49-89-32356-0 D- 80807 Muenchen Fax : +49-89-32356-234
hello El 04/10/2006, a las 10:19, Gert Doering escribió:
I'd propose a minimum AW of /24, to be raised to a /22 if the LIR has sent one? two? person(s) to a LIR training course (and the trainer is convinced that these people have listened and understood the rules).
Considering that the training course doesn't have an examn, it's dificult for the trainer to know how much people has understood. I would modify your propossal to: 1st AW or /24 Raised AW to /22 after the first or second allocation is made if there are no errors in those. Regards, Fernando ------------------------------------------------ Fernando Garcia |Tel: +34 91 4359687 EUROCOMERCIAL I&C SA |Fax: +34 91 4313240 Valentín Beato, 5 |e-mail: fgarcia@eurocomercial.es E-28037 Madrid | Spain |http://www.eurocomercial.es
Hi, On Wed, Oct 04, 2006 at 10:25:50AM +0200, Fernando García wrote:
I would modify your propossal to: 1st AW or /24 Raised AW to /22 after the first or second allocation is made if there are no errors in those.
Well, that's the way it is now - if you send a few requests to the NCC hostmasters, and they are approved, you're granted an AW of the size of these requests (roughly so). So we wouldn't actually need the second sentence, it would account to "minimum AW of a /24"... - which I could live with, but Leo is clearly aiming for "more". Gert Doering -- NetMaster -- Total number of prefixes smaller than registry allocations: 98999 SpaceNet AG Mail: netmaster@Space.Net Joseph-Dollinger-Bogen 14 Tel : +49-89-32356-0 D- 80807 Muenchen Fax : +49-89-32356-234
Hi, Gert Doering wrote:
Hi,
On Wed, Oct 04, 2006 at 10:25:50AM +0200, Fernando García wrote:
I would modify your propossal to: 1st AW or /24 Raised AW to /22 after the first or second allocation is made if there are no errors in those.
Well, that's the way it is now - if you send a few requests to the NCC hostmasters, and they are approved, you're granted an AW of the size of these requests (roughly so).
So we wouldn't actually need the second sentence, it would account to "minimum AW of a /24"... - which I could live with, but Leo is clearly aiming for "more".
For (standard) IPv6 assignments, you don't even have to ask RIPE in most cases by default. So, raising the IPv4 AW and thus (supposingly) reducing the workload on RIPE Hostmasters could probably leave to less cost and lower RIPE Membership fees - and/or better services as pointed out in the proposal. Although i'm a bit puzzled about this proposal (never thought about changing the AW handling lately), i'm supporting the proposal in general (larger AWs). I don't think there will be much negative impact. One might discuss the details though. I wouldn't object keeping a slow-start mechanism (to proof some kind of clue). Probably starting with - whatever - /24? and doubling the AW every two or three (RIPE approved) requests, regardless of their size. But this probably doesn't make that much sense either. -- ======================================================================== = Sascha Lenz SLZ-RIPE slz@baycix.de = = Network Operations = = BayCIX GmbH, Landshut * PGP public Key on demand * = ========================================================================
Sascha Lenz wrote: [...]
For (standard) IPv6 assignments, you don't even have to ask RIPE in most cases by default. [...]
Which makes me remember again that I added the following comment to similar discussions in the past: "I don't believe that the /`whatever` is a useful yardstick to reasonably measure the complexity, difficulty or quality of request assessment for an assignment." My only problem is that I don't know which conclusion to draw from this, regarding the AW... Wilfried.
Fernando García wrote: [...]
Considering that the training course doesn't have an examn, it's dificult for the trainer to know how much people has understood.
...and there's https://e-learning.ripe.net/ :-) Wilfried.
participants (5)
-
Fernando García -
Filiz Yilmaz -
Gert Doering -
Sascha Lenz -
Wilfried Woeber, UniVie/ACOnet