Re: [address-policy-wg] New Policy Proposal (PI - PA Transfer)
RIPE 67 is ended. Just wondering if you have any updates?
We will be addressing this as part of our "Feedback From RIPE NCC Registration Services" presentation at RIPE 67. Based on the feedback we receive from here, we will work with the Working Group Chairs on a way forward. Kind regards, Andrea Cima RIPE NCC
Think I'll just "hijack" this discussion ;p I fully support the idea of uniting PI and PA into one policy, it's just numbers anyway. However, the presentation on the topic from the three (just two onsite) volunteers was very interesting, the consequences and how much it affected was amazing. How did we dig us into such a mess? With that background I don't think we can continue forward on the started track, the impact everywhere and amount of work are too big. I think we should take two step back and cleanup the mess inside the PI "domain", try to clean and polish it, link it toward reality, all the things that are connected and linked. Not so much about changing alot, just clean it up. Or we can fix what's obvious broken along the way of course. We should also do the same on the PA side, but probably not that much there to clean up. And this should be done both for v4 and v6 before we again in a year or two try to unite it into one great policy. anyhow, as long as we don't drag this mess over into v6 the future isn't all that bad:-) just my 0,02 NOK :-) --- Roger J --- On Mon, Oct 21, 2013 at 1:42 PM, Dmitriy Zemlyanoy <dmitriy@deltahost.com.ua> wrote:
RIPE 67 is ended. Just wondering if you have any updates?
We will be addressing this as part of our "Feedback From RIPE NCC Registration Services" presentation at RIPE 67. Based on the feedback we receive from here, we will work with the Working Group Chairs on a way forward. Kind regards, Andrea Cima RIPE NCC
-- Roger Jorgensen | ROJO9-RIPE rogerj@gmail.com | - IPv6 is The Key! http://www.jorgensen.no | roger@jorgensen.no
Hi, On Mon, Oct 21, 2013 at 02:05:14PM +0200, Roger Jørgensen wrote:
Think I'll just "hijack" this discussion ;p
I fully support the idea of uniting PI and PA into one policy, it's just numbers anyway.
Uh, you're confusing two issues. *This* thread was about being able to turn IPv*4* PI into PA, while keeping the distinction. Andrea Cima presented at the RIPE meeting, and I expect to see a summary and "next steps" from him soon - especially for the bits that are missing policywise, and Erik Bais volunteered to work with us to fix those.
However, the presentation on the topic from the three (just two onsite) volunteers was very interesting, the consequences and how much it affected was amazing. How did we dig us into such a mess?
*That* is actually about IPv6 land (PA/PI unification, only having a single "type of addresses" anymore), and I'll open a separate thread on that "soonish". Gert Doering -- APWG chair -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
On Mon, Oct 21, 2013 at 2:08 PM, Gert Doering <gert@space.net> wrote:
Hi,
On Mon, Oct 21, 2013 at 02:05:14PM +0200, Roger Jørgensen wrote:
Think I'll just "hijack" this discussion ;p
I fully support the idea of uniting PI and PA into one policy, it's just numbers anyway.
Uh, you're confusing two issues. *This* thread was about being able to turn IPv*4* PI into PA, while keeping the distinction.
Andrea Cima presented at the RIPE meeting, and I expect to see a summary and "next steps" from him soon - especially for the bits that are missing policywise, and Erik Bais volunteered to work with us to fix those.
oh, see that now, sorry.
However, the presentation on the topic from the three (just two onsite) volunteers was very interesting, the consequences and how much it affected was amazing. How did we dig us into such a mess?
*That* is actually about IPv6 land (PA/PI unification, only having a single "type of addresses" anymore), and I'll open a separate thread on that "soonish".
'ki, we're waiting:-) -- Roger Jorgensen | ROJO9-RIPE rogerj@gmail.com | - IPv6 is The Key! http://www.jorgensen.no | roger@jorgensen.no
Hi, small clarification... On Mon, Oct 21, 2013 at 02:08:45PM +0200, Gert Doering wrote:
On Mon, Oct 21, 2013 at 02:05:14PM +0200, Roger Jørgensen wrote:
Think I'll just "hijack" this discussion ;p
I fully support the idea of uniting PI and PA into one policy, it's just numbers anyway.
Uh, you're confusing two issues. *This* thread was about being able to turn IPv*4* PI into PA, while keeping the distinction.
Andrea Cima presented at the RIPE meeting, and I expect to see a summary and "next steps" from him soon - especially for the bits that are missing policywise, and Erik Bais volunteered to work with us to fix those.
This is where *I* was confused about things - Erik Bais volunteered, but not to fix this particular bit about IPv4 PI, but to work out something about IPv4 PI transfers. Sorted out by private mail in the meantime. But I think we have found another volunteer for the minimum allocation size obstacle regarding IPv4 PA->PI conversion. Also related to IPv4 PI, but differently... (And I can already hear Randy Bush tell me that our way of thinking is too complicated... maybe it is, but we're such orderly folks here... we can do this all by policy, without lawyers :-) ) Gert Doering -- APWG chair -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
Gert, you still seem to be a bit stressed out - no relaxing weekend?! ;-) On 21.10.2013 20:40, Gert Doering wrote:
But I think we have found another volunteer for the minimum allocation size obstacle regarding IPv4 PA->PI conversion. Also related to IPv4 PI, but differently...
This should read "regarding IPv4 PI->PA conversion" - as it concerns converting "ASSINGED PI" IPv4 space of LIRs into "ALLOCATED PA" space for the respective LIR. The (main) question here is: what to do with PI space that is smaller than the minimum allocation size. Cheers, -C.
Hi, On Mon, Oct 21, 2013 at 09:35:17PM +0200, Carsten Schiefner wrote:
you still seem to be a bit stressed out - no relaxing weekend?! ;-)
Not that much... conflicting appointments on the weekend right after the RIPE meeting... I thought I had enough sleep, but seems not.
On 21.10.2013 20:40, Gert Doering wrote:
But I think we have found another volunteer for the minimum allocation size obstacle regarding IPv4 PA->PI conversion. Also related to IPv4 PI, but differently...
This should read "regarding IPv4 PI->PA conversion" - as it concerns converting "ASSINGED PI" IPv4 space of LIRs into "ALLOCATED PA" space for the respective LIR.
Yeah, of course. Sorry.
The (main) question here is: what to do with PI space that is smaller than the minimum allocation size.
"Just permitting the conversion" ("in one piece, not to be fragmented further") would be one option :-) - no extra routing table impact, no extra address space consumption, and possibly improved documentation. But that's not for me to decide, it's for you to propose and for the WG to decide :-) (but it matches what I read here). Gert Doering -- APWG chair -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
Hi to everybody I don't understand what is the problem to return the possibility to turn PI to PA? If it was somedays = that means that is possible. What troubles RIPE are expecting to see - if they are still oscillate with the solution? IPv6 unfortunately is not using as It was expecting to be in use. Good if 5% of providers are using IPv6. Now almost all are still using IPv4. The principles in this policy generates a lot of troubles. Some people are starting to steal or use somebodies networks illegally. Don't close your eyes. You know how they can do this. What should all other companies do if they can't get their own addresses? They will ask for rent. But the price now is too high. And will not go down if something will not changes. As I see the solution to soften the migration from v4 to v6 - this solution is a flexible POLICY. Let's return the possibility to transfer PI to PA and ease the life of a lot of small companies which need their own addresses. What to do with PI space that is smaller than the minimum allocation size? I see excellent way - Make a minimum /22 and just wait for a little bit - when addresses will end finally. Than every block can be turned into the PA. Best regards, Kseniya 21.10.2013 22:35, Carsten Schiefner пишет:
Gert,
you still seem to be a bit stressed out - no relaxing weekend?! ;-)
On 21.10.2013 20:40, Gert Doering wrote:
But I think we have found another volunteer for the minimum allocation size obstacle regarding IPv4 PA->PI conversion. Also related to IPv4 PI, but differently...
This should read "regarding IPv4 PI->PA conversion" - as it concerns converting "ASSINGED PI" IPv4 space of LIRs into "ALLOCATED PA" space for the respective LIR.
The (main) question here is: what to do with PI space that is smaller than the minimum allocation size.
Cheers,
-C.
-- С ув. Ксения Сокол
Exactly. I also think that current (disallowing) transfer practice for block /22 (and larger) - is an artificial problem. I think there is no problem to allow transfer PI->PA for blocks /22 (and larger) right now. And after that we can continue discussing about blocks less then /22. It seems that question about transfering blocks less then /22 is not easy, but there is no question for blocks /22 (and large). Delaying this for discussing question about blocks less then /22 - we making situation more and more hard each day. Agiotage around IPv4 is growing each day. So, let's make some cool down of situation - let's allow transfer /22 (and larger) right now and leave the question of less then /22 for discussion. It will improve situation right now and give us some time to discuss about blocks less then /22 -- Dmitriy Zemlyanoy. DeltaHost. http://deltahost.com ksyu@netassist.ua писал 2013-10-29 20:47:
Hi to everybody
I don't understand what is the problem to return the possibility to turn PI to PA? If it was somedays = that means that is possible. What troubles RIPE are expecting to see - if they are still oscillate with the solution?
IPv6 unfortunately is not using as It was expecting to be in use. Good if 5% of providers are using IPv6.
Now almost all are still using IPv4.
The principles in this policy generates a lot of troubles. Some people are starting to steal or use somebodies networks illegally. Don't close your eyes. You know how they can do this.
What should all other companies do if they can't get their own addresses? They will ask for rent. But the price now is too high. And will not go down if something will not changes.
As I see the solution to soften the migration from v4 to v6 - this solution is a flexible POLICY.
Let's return the possibility to transfer PI to PA and ease the life of a lot of small companies which need their own addresses.
What to do with PI space that is smaller than the minimum allocation size? I see excellent way - Make a minimum /22 and just wait for a little bit - when addresses will end finally. Than every block can be turned into the PA.
Best regards, Kseniya
21.10.2013 22:35, Carsten Schiefner пишет:
Gert,
you still seem to be a bit stressed out - no relaxing weekend?! ;-)
On 21.10.2013 20:40, Gert Doering wrote:
But I think we have found another volunteer for the minimum allocation size obstacle regarding IPv4 PA->PI conversion. Also related to IPv4 PI, but differently...
This should read "regarding IPv4 PI->PA conversion" - as it concerns converting "ASSINGED PI" IPv4 space of LIRs into "ALLOCATED PA" space for the respective LIR.
The (main) question here is: what to do with PI space that is smaller than the minimum allocation size.
Cheers,
-C.
On Tue, Oct 29, 2013 at 7:47 PM, ksyu@netassist.ua <ksyu@netassist.ua>wrote:
IPv6 unfortunately is not using as It was expecting to be in use. Good if 5% of providers are using IPv6.
Now almost all are still using IPv4.
This is not unexpected.
From my point of view, IPv6 uptake is higher than what I expected.
That has only just begun, and I honestly don't think low uptake of IPv6 (if it was true!) is a good argument for policy changes in IPv4. -- Jan
participants (6)
-
Carsten Schiefner -
Dmitriy Zemlyanoy -
Gert Doering -
Jan Ingvoldstad -
ksyu＠netassist.ua -
Roger Jørgensen