Re: [address-policy-wg] 2011-04 Discussion Period extended until 30 January (Extension of the Minimum Size for IPv6 Initial Allocation)
On 02/01/2012 14:10, Emilio Madaio wrote:
- The proposed new section 5.1.2 was reworded - Section 5.7 was not removed but it was reworded
two issues here: 1. I don't agree with this revised version for the reasons outlined in:
https://www.ripe.net/ripe/mail/archives/address-policy-wg/2011-November/0065...
specifically, there is still no justification required to move from /32 to /29. 2. if the plan is to use the entire /29 for the purposes of 6rd (or other transition tech) - so that you can assign up to a /62 for each 6rd end-user, then what address space does the LIR use for the rest of its allocation requirements? I.e. will another allocation will be required for traditional ipv6 assignments? Nick
On 1/4/12 8:01 PM, Nick Hilliard wrote:
On 02/01/2012 14:10, Emilio Madaio wrote:
- The proposed new section 5.1.2 was reworded - Section 5.7 was not removed but it was reworded
two issues here:
1. I don't agree with this revised version for the reasons outlined in:
https://www.ripe.net/ripe/mail/archives/address-policy-wg/2011-November/0065...
specifically, there is still no justification required to move from /32 to /29.
Nick, hi. Yes, as you said. We have been through this discussion and there is no point in re-doing it again. Your suggestion is just inserting the pointless obstacle, making LIRs claim they will do 6rd even if they do not intend to do so - to get /29. We can go around in circles, but I'm not sure we need this :) We thought to insert partially your idea with suggestion, that LIR should clarify just for documentation purposes, why they need more than /29, but at the end decided, that this is not adding anything, just making the policy longer.
2. if the plan is to use the entire /29 for the purposes of 6rd (or other transition tech) - so that you can assign up to a /62 for each 6rd end-user, then what address space does the LIR use for the rest of its allocation requirements? I.e. will another allocation will be required for traditional ipv6 assignments?
30 + 32 = 62 :) Cheers, Jan
On 1/5/12 12:58 AM, Jan Zorz @ go6.si wrote:
We thought to insert partially your idea with suggestion, that LIR should clarify just for documentation purposes, why they need more than /29, but at the end decided, that this is not adding anything, just making the policy longer.
Sorry, typo, it should say "why they need more than /32" Cheers, Jan
On 04/01/2012 23:58, Jan Zorz @ go6.si wrote:
Yes, as you said. We have been through this discussion and there is no point in re-doing it again. Your suggestion is just inserting the pointless obstacle, making LIRs claim they will do 6rd even if they do not intend to do so - to get /29. We can go around in circles, but I'm not sure we need this :)
We thought to insert partially your idea with suggestion, that LIR should clarify just for documentation purposes, why they need more than /29, but at the end decided, that this is not adding anything, just making the policy longer.
Hi Jan, maybe we need to disagree. I don't support the proposal as-is, but would support the proposal if it were to include minimal justification for /29 (based on the current default of /32). To recap, the reason I hold this view is: - it is a minimal change which requires virtually no overhead by the LIR, but will get them to think about whether they really need the space or not. - many LIRs will never need to use 6rd or any other transition technology, so assigning an extra 3 bits of address space is wasteful - For the sort of LIR which doesn't require a transition technology like this, /32 is probably a lot more than the LIR will ever need anyway. - RIPE and the RIPE NCC have a duty of good stewardship to the resources which they maintain. Increasing the amount of space allocated to LIRS by a factor of 8 without any justification whatever is (imo) bad stewardship of resources. Nick
On 1/9/12 5:23 PM, Nick Hilliard wrote:
maybe we need to disagree. I don't support the proposal as-is, but would support the proposal if it were to include minimal justification for /29 (based on the current default of /32).
Nick, hi. Ok, I hear what you are saying.
To recap, the reason I hold this view is:
- it is a minimal change which requires virtually no overhead by the LIR, but will get them to think about whether they really need the space or not.
I think the price difference in LIR membership will probably make them think - more resources, bigger LIR ;)
- many LIRs will never need to use 6rd or any other transition technology, so assigning an extra 3 bits of address space is wasteful
- For the sort of LIR which doesn't require a transition technology like this, /32 is probably a lot more than the LIR will ever need anyway.
- RIPE and the RIPE NCC have a duty of good stewardship to the resources which they maintain. Increasing the amount of space allocated to LIRS by a factor of 8 without any justification whatever is (imo) bad stewardship of resources.
Let's see what others think - WG, any thoughts? Cheers, Jan
Am 10.01.2012 09:28, schrieb Jan Zorz @ go6.si:
- it is a minimal change which requires virtually no overhead by the LIR, but will get them to think about whether they really need the space or not.
I think the price difference in LIR membership will probably make them think - more resources, bigger LIR ;)
Perhaps we should include a footnode a /29 may increase the membership fee ;-) I don't think a minimal justification gets us any benefit. If a LIR just has to say '6RD' he can do this without any thinking. Let's skip this step.
- many LIRs will never need to use 6rd or any other transition technology, so assigning an extra 3 bits of address space is wasteful
- For the sort of LIR which doesn't require a transition technology like this, /32 is probably a lot more than the LIR will ever need anyway.
We don't intend to use 6RD but we would like to extend our allocation to /29 in order to configure proper v6 address-pools on our access-routers. For us the proposal will resolve a real existing problem. If we wouldn't need a /29 I wouldn't request it. Why should I? I'm not afraid not getting more v6 space any time later.
- RIPE and the RIPE NCC have a duty of good stewardship to the resources which they maintain. Increasing the amount of space allocated to LIRS by a factor of 8 without any justification whatever is (imo) bad stewardship of resources.
Let's see what others think - WG, any thoughts?
For me good stewardship also means to take care about the problems of it's members. Sometimes it may be difficult to find the right balance between not beeing wasteful and allocating the needed ip space. But as this workgroup is listening to the different pro's and con's it's already doing a good job. I see this a good stewardship. And I support the proposal as it is. cheers, Michael -- Michael Adams Tel: +49 221 2222 657 Network Engineering & Design Fax: +49 221 2222 7657 NetCologne Geschäftsführer Gesellschaft für Telekommunikation mbH Dr. Hans Konle (Sprecher) Am Coloneum 9 Dipl.-Ing. Karl-Heinz Zankel 50829 Köln HRB 25580, Amtsgericht Köln
Hi, I opt for extending allocation to /29 for these LIRs which simply request it. Kind regards, Jurek
On Wed, Jan 11, 2012 at 02:13:25PM +0100, Michael Adams wrote:
- For the sort of LIR which doesn't require a transition technology like this, /32 is probably a lot more than the LIR will ever need anyway.
We don't intend to use 6RD but we would like to extend our allocation to /29 in order to configure proper v6 address-pools on our access-routers. For us the proposal will resolve a real existing problem. If we wouldn't need a /29 I wouldn't request it. Why should I? I'm not afraid not getting more v6 space any time later.
Still, having the "initial request" policy being severely more relaxed than the "additional request" policy poses looming problems for those who will have to request more address space later on. But it's a step into the right direction... Best regards, Daniel (sharing the same pain) -- CLUE-RIPE -- Jabber: dr@cluenet.de -- dr@IRCnet -- PGP: 0xA85C8AA0
Hi Jan, On Tue, Jan 10, 2012 at 09:28:42AM +0100, Jan Zorz @ go6.si wrote:
On 1/9/12 5:23 PM, Nick Hilliard wrote:
maybe we need to disagree. I don't support the proposal as-is, but would support the proposal if it were to include minimal justification for /29 (based on the current default of /32).
- it is a minimal change which requires virtually no overhead by the LIR, but will get them to think about whether they really need the space or not.
I think the price difference in LIR membership will probably make them think - more resources, bigger LIR ;)
That's something for a requester to think about. The minimal change Nick talks about above and tracking the reason of usage is also of possible interest to the rest of the community. Well it rather seems like a minority of it so far. The proposed policy change reads like this: No questions asked if the requested IPv6 address space size (to be) allocated to a LIR is >=/32 && <=/29 whatever the intended usage reason is, e.g. the LIR's customer base size, a need for a transition technology. Have I missed anything ? Martin
On 1/11/12 3:42 PM, Martin Stanislav wrote:
I think the price difference in LIR membership will probably make them think - more resources, bigger LIR ;)
That's something for a requester to think about.
Hi, Exactly. It's up to how we define charging schema, but basically, yes, you are right.
The minimal change Nick talks about above and tracking the reason of usage is also of possible interest to the rest of the community. Well it rather seems like a minority of it so far.
The proposed policy change reads like this: No questions asked if the requested IPv6 address space size (to be) allocated to a LIR is>=/32&& <=/29 whatever the intended usage reason is, e.g. the LIR's customer base size, a need for a transition technology. Have I missed anything ?
No. That's correct understanding. :) Cheers, Jan
Hi, Jan Zorz wrote: [...]
I think the price difference in LIR membership will probably make them think - more resources, bigger LIR ;)
As the WG does not control the charging scheme it does not seem sensible to rely on the charging scheme remaining the same for the effectiveness of a policy created in the WG. Regards, Leo
On 1/12/12 5:22 PM, Leo Vegoda wrote:
I think the price difference in LIR membership will probably make them think - more resources, bigger LIR ;)
As the WG does not control the charging scheme it does not seem sensible to rely on the charging scheme remaining the same for the effectiveness of a policy created in the WG.
Leo, hi thnx for comment. It is not a matter of "relying", as we don't control charging scheme - and charging scheme must not affect the policy itself and vice-versa. Agree fully, but currently I can not see that this affects the proposal - but we need to keep this in mind in the future. Cheers, Jan
participants (7)
-
Daniel Roesen -
Jan Zorz ＠ go6.si -
Jerzy Pawlus -
Leo Vegoda -
Martin Stanislav -
Michael Adams -
Nick Hilliard