Keyboard Shortcuts
Thread View
j: Next unread messagek: Previous unread messagej a: Jump to all threadsj l: Jump to MailingList overview
You're probably referring to Cisco's early BGP implementation in IOS where the BGP Scanner process ran every 60 seconds (although it could be tuned to run at much shorter intervals).
Software has come a very long way since then, Ramesh, and IOS, in particular, has benefited quite a bit. Gone is the BGP Scanner, which has since been replaced by Next Hop Address Tracking. Next Hop Address Tracking has also been superseded by PIC (Prefix Independent Convergence).
All these improvements in software have meant convergence times for BGP have improved greatly, and the 60-second crunch routers have to do is no longer necessary provided they are running the software that supports these new features.
Hi Mark, agree, there is lot of improvement in SW coding. Cisco has also improved in XOR release. However, number of routers running XOR is very low. Cisco as market leader runs same process in IOS used in most of routers at present. Same is with other vendors and impact possibly may be different. There is no intention of summarizing routers available in different AS.
regards Ramesh Chandra | GM ? Network Engineering | NSG | Bharti Airtel Limited | Phone# +91 124 4243897, 9810300704 ----- Forwarded by Ramesh Chandra/India/Airtel on 08/22/2011 10:56 AM -----
Mark Tinka mtinka@globaltransit.net 08/19/2011 11:17 PM Please respond to mtinka@globaltransit.net
To sig-policy@lists.apnic.net cc ramesh.chandra@airtel.in Subject Re: [sig-policy] Pro-100:National IP Address Plan - Allocation of country-wide IP address blocks P Please do not print this e-mail unless it is absolutely necessary P
On Friday, August 19, 2011 09:54:04 PM ramesh.chandra@airtel.in wrote:
Appreciate your query and let us look at simple Engineering example for a Enterprise house. One Enterprise customer manage 10 IP pools allocated to him as per his business demand in previous years. He is announcing these 10 routes in BGP to upstream carriers, NAP/NIXI and other private peering he has with others. Every connected BGP listner has to process every 60 seconds these 10 routes and install in routing table and this continues to thousands of routers in Internet.
You're probably referring to Cisco's early BGP implementation in IOS where the BGP Scanner process ran every 60 seconds (although it could be tuned to run at much shorter intervals).
Software has come a very long way since then, Ramesh, and IOS, in particular, has benefited quite a bit. Gone is the BGP Scanner, which has since been replaced by Next Hop Address Tracking. Next Hop Address Tracking has also been superseded by PIC (Prefix Independent Convergence).
All these improvements in software have meant convergence times for BGP have improved greatly, and the 60-second crunch routers have to do is no longer necessary provided they are running the software that supports these new features.
If these 10 pools become one continuous block then these thousands of routers shall scan and install one route instead 10 routes in previous case under that AS. This helps to reduce few CPU cycles and these CPU cycles become significant when we talk thousand routers repeating the same.
Unless you're talking about route crunching within a single AS, I'd be more concerned about one router processing thousands of routes, as opposed to a thousand routers - each in a different AS - processing a handful of routes.
Lesser routes in BGP required less time to install and hence faster Convergence time. Presently, convergence time for 10k IPv4/Ipv6 routes takes approx 1 second. Convergence time shall reduce significantly when everyone summarise their own routes before announcing.
Agree, but do you think it is feasible that an entire country, made up of several ISP's, can announce a single IPv6 allocation to the Internet? Serious question.
Mark.
This e-mail and any files transmitted with it are for the sole use of the intended recipient(s) and may contain confidential and privileged information. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies and the original message. Any unauthorized review, use, disclosure,dissemination, forwarding, printing or copying of this email or any action taken in reliance on this e-mail is strictly prohibited and may be unlawful. The recipient acknowledges that Bharti Airtel Limited or its subsidiaries and associated companies(collectively "Bharti Airtel Limited"),are unable to exercise control or ensure or guarantee the integrity of/overthe contents of the information contained in e-mail transmissions and further acknowledges that any views expressed in this message are those of the individual sender and no binding nature of the message shall be implied or assumed unless the sender does so expressly with due authority of Bharti Airtel Limited. Before opening any attachments please check them for viruses and defects.
More so, next hop address tracking is in the transition stage and not the ecosystem. The prop 100 is for NIRs and shud be debated in accordance only.
Regards and best wishes,
Naresh Ajwani Sent from my iPad
On Aug 22, 2011, at 11:15, ramesh.chandra@airtel.in wrote:
You're probably referring to Cisco's early BGP implementation in IOS where the BGP Scanner process ran every 60 seconds (although it could be tuned to run at much shorter intervals).
Software has come a very long way since then, Ramesh, and IOS, in particular, has benefited quite a bit. Gone is the BGP Scanner, which has since been replaced by Next Hop Address Tracking. Next Hop Address Tracking has also been superseded by PIC (Prefix Independent Convergence).
All these improvements in software have meant convergence times for BGP have improved greatly, and the 60-second crunch routers have to do is no longer necessary provided they are running the software that supports these new features.
Hi Mark, agree, there is lot of improvement in SW coding. Cisco has also improved in XOR release. However, number of routers running XOR is very low. Cisco as market leader runs same process in IOS used in most of routers at present. Same is with other vendors and impact possibly may be different. There is no intention of summarizing routers available in different AS.
regards Ramesh Chandra | GM – Network Engineering | NSG | Bharti Airtel Limited | Phone# +91 124 4243897, 9810300704 ----- Forwarded by Ramesh Chandra/India/Airtel on 08/22/2011 10:56 AM -----
Mark Tinka mtinka@globaltransit.net 08/19/2011 11:17 PM
Please respond to mtinka@globaltransit.net
To sig-policy@lists.apnic.net cc ramesh.chandra@airtel.in Subject Re: [sig-policy] Pro-100:National IP Address Plan - Allocation of country-wide IP address blocks
P Please do not print this e-mail unless it is absolutely necessary POn Friday, August 19, 2011 09:54:04 PM ramesh.chandra@airtel.in wrote:
Appreciate your query and let us look at simple Engineering example for a Enterprise house. One Enterprise customer manage 10 IP pools allocated to him as per his business demand in previous years. He is announcing these 10 routes in BGP to upstream carriers, NAP/NIXI and other private peering he has with others. Every connected BGP listner has to process every 60 seconds these 10 routes and install in routing table and this continues to thousands of routers in Internet.
You're probably referring to Cisco's early BGP implementation in IOS where the BGP Scanner process ran every 60 seconds (although it could be tuned to run at much shorter intervals).
Software has come a very long way since then, Ramesh, and IOS, in particular, has benefited quite a bit. Gone is the BGP Scanner, which has since been replaced by Next Hop Address Tracking. Next Hop Address Tracking has also been superseded by PIC (Prefix Independent Convergence).
All these improvements in software have meant convergence times for BGP have improved greatly, and the 60-second crunch routers have to do is no longer necessary provided they are running the software that supports these new features.
If these 10 pools become one continuous block then these thousands of routers shall scan and install one route instead 10 routes in previous case under that AS. This helps to reduce few CPU cycles and these CPU cycles become significant when we talk thousand routers repeating the same.
Unless you're talking about route crunching within a single AS, I'd be more concerned about one router processing thousands of routes, as opposed to a thousand routers - each in a different AS - processing a handful of routes.
Lesser routes in BGP required less time to install and hence faster Convergence time. Presently, convergence time for 10k IPv4/Ipv6 routes takes approx 1 second. Convergence time shall reduce significantly when everyone summarise their own routes before announcing.
Agree, but do you think it is feasible that an entire country, made up of several ISP's, can announce a single IPv6 allocation to the Internet? Serious question.
Mark. This e-mail and any files transmitted with it are for the sole use of the intended recipient(s) and may contain confidential and privileged information. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies and the original message. Any unauthorized review, use, disclosure,dissemination, forwarding, printing or copying of this email or any action taken in reliance on this e-mail is strictly prohibited and may be unlawful. The recipient acknowledges that Bharti Airtel Limited or its subsidiaries and associated companies(collectively "Bharti Airtel Limited"),are unable to exercise control or ensure or guarantee the integrity of/overthe contents of the information contained in e-mail transmissions and further acknowledges that any views expressed in this message are those of the individual sender and no binding nature of the message shall be implied or assumed unless the sender does so expressly with due authority of Bharti Airtel Limited. Before opening any attachments please check them for viruses and defects.
<signature.asc>
sig-policy: APNIC SIG on resource management policy *
sig-policy mailing list sig-policy@lists.apnic.net http://mailman.apnic.net/mailman/listinfo/sig-policy
Dear All
I have seen lot of comments and debate for Prop.100 and from some of the comments , it is clear some of the members are not clear about the very purpose of the proposal , so it is requested to all to kindly go through our original proposal and clarifications issued on 17/8/2011 then it will be clear to them that this proposal is not talking only of some specific countries, but taking care of all economies in the APNIC Region, as we have said in our original proposal that country wise Address block can be allocated to NIRs( For those economies those who are having NIR) and further allocation to Different organizations can be done by the NIR and of course through APNIC, keeping alignment with APNIC policy and for non NIR countries this big block can be reserved in APNIC secretariat and allocations to different organizations in that country can be done out of this big block.
In this process all economies will be taken care of and will not be left out like in case of IPv4 era, while maintaining the control of RIRs ,This was the theme of the whole proposal . Many of the reputed members started discussions on the technicalities of the advantages, as quoted in the proposal and opposed the proposal on these grounds , It is suggested to all APNIC community members to see the theme and spirit of the proposal as explained above and work towards overall interest of all countries and organizations in APNIC region.
Regards
R M AGARWAL
DDG (NT),
Department of Telecommunications,
Room no. 1104, Sanchar Bhawan
20, Ashoka Road, New Delhi - 110001
Mob: +91- 9868133440
Off no. +91 11 23372606
.
From: sig-policy-bounces@lists.apnic.net [mailto:sig-policy-bounces@lists.apnic.net] On Behalf Of Naresh Ajwani Sent: 22 August 2011 12:02 To: ramesh.chandra@airtel.in Cc: sig-policy@lists.apnic.net Subject: Re: [sig-policy] Fw: Pro-100:National IP Address Plan - Allocation of country-wide IP address blocks
And
More so, next hop address tracking is in the transition stage and not the ecosystem. The prop 100 is for NIRs and shud be debated in accordance only.
Regards and best wishes,
Naresh Ajwani
Sent from my iPad
On Aug 22, 2011, at 11:15, ramesh.chandra@airtel.in wrote:
You're probably referring to Cisco's early BGP implementation in IOS where the BGP Scanner process ran every 60 seconds (although it could be tuned to run at much shorter intervals).
Software has come a very long way since then, Ramesh, and IOS, in particular, has benefited quite a bit. Gone is the BGP Scanner, which has since been replaced by Next Hop Address Tracking. Next Hop Address Tracking has also been superseded by PIC (Prefix Independent Convergence).
All these improvements in software have meant convergence times for BGP have improved greatly, and the 60-second crunch routers have to do is no longer necessary provided they are running the software that supports these new features.
Hi Mark, agree, there is lot of improvement in SW coding. Cisco has also improved in XOR release. However, number of routers running XOR is very low. Cisco as market leader runs same process in IOS used in most of routers at present. Same is with other vendors and impact possibly may be different. There is no intention of summarizing routers available in different AS.
regards Ramesh Chandra | GM – Network Engineering | NSG | Bharti Airtel Limited | Phone# +91 124 4243897, 9810300704 ----- Forwarded by Ramesh Chandra/India/Airtel on 08/22/2011 10:56 AM -----
Mark Tinka mtinka@globaltransit.net
08/19/2011 11:17 PM
Please respond to mtinka@globaltransit.net
To
sig-policy@lists.apnic.net
cc
ramesh.chandra@airtel.in
Subject
Re: [sig-policy] Pro-100:National IP Address Plan - Allocation of country-wide IP address blocks
P Please do not print this e-mail unless it is absolutely necessary P
On Friday, August 19, 2011 09:54:04 PM ramesh.chandra@airtel.in wrote:
Appreciate your query and let us look at simple Engineering example for a Enterprise house. One Enterprise customer manage 10 IP pools allocated to him as per his business demand in previous years. He is announcing these 10 routes in BGP to upstream carriers, NAP/NIXI and other private peering he has with others. Every connected BGP listner has to process every 60 seconds these 10 routes and install in routing table and this continues to thousands of routers in Internet.
You're probably referring to Cisco's early BGP implementation in IOS where the BGP Scanner process ran every 60 seconds (although it could be tuned to run at much shorter intervals).
Software has come a very long way since then, Ramesh, and IOS, in particular, has benefited quite a bit. Gone is the BGP Scanner, which has since been replaced by Next Hop Address Tracking. Next Hop Address Tracking has also been superseded by PIC (Prefix Independent Convergence).
All these improvements in software have meant convergence times for BGP have improved greatly, and the 60-second crunch routers have to do is no longer necessary provided they are running the software that supports these new features.
If these 10 pools become one continuous block then these thousands of routers shall scan and install one route instead 10 routes in previous case under that AS. This helps to reduce few CPU cycles and these CPU cycles become significant when we talk thousand routers repeating the same.
Unless you're talking about route crunching within a single AS, I'd be more concerned about one router processing thousands of routes, as opposed to a thousand routers - each in a different AS - processing a handful of routes.
Lesser routes in BGP required less time to install and hence faster Convergence time. Presently, convergence time for 10k IPv4/Ipv6 routes takes approx 1 second. Convergence time shall reduce significantly when everyone summarise their own routes before announcing.
Agree, but do you think it is feasible that an entire country, made up of several ISP's, can announce a single IPv6 allocation to the Internet? Serious question.
Mark.
This e-mail and any files transmitted with it are for the sole use of the intended recipient(s) and may contain confidential and privileged information. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies and the original message. Any unauthorized review, use, disclosure,dissemination, forwarding, printing or copying of this email or any action taken in reliance on this e-mail is strictly prohibited and may be unlawful. The recipient acknowledges that Bharti Airtel Limited or its subsidiaries and associated companies(collectively "Bharti Airtel Limited"),are unable to exercise control or ensure or guarantee the integrity of/overthe contents of the information contained in e-mail transmissions and further acknowledges that any views expressed in this message are those of the individual sender and no binding nature of the message shall be implied or assumed unless the sender does so expressly with due authority of Bharti Airtel Limited. Before opening any attachments please check them for viruses and defects.
<signature.asc>
* sig-policy: APNIC SIG on resource management policy * _______________________________________________ sig-policy mailing list sig-policy@lists.apnic.net http://mailman.apnic.net/mailman/listinfo/sig-policy
On 22/08/2011, at 7:52 PM, RAKESH MOHAN AGARWAL wrote: In this process all economies will be taken care of and will not be left out like in case of IPv4 era,
If your policy proposal had been adopted in, say, 1996, in what ways would the resulting allocation of IPv4 resources differ from what we have now?
- mark
-- Mark Newton Email: newton@internode.com.aumailto:newton@internode.com.au (W) Network Engineer Email: newton@atdot.dotat.orgmailto:newton@atdot.dotat.org (H) Internode Pty Ltd Desk: +61-8-82282999 "Network Man" - Anagram of "Mark Newton" Mobile: +61-416-202-223
Rakish jee, let's limit our clarification to v6 only.
RIRs receive v6 blocks in /12 units vis-a-vis to /8 for v4.
Regards and best wishes,
Naresh Ajwani Sent from my iPad
On Aug 22, 2011, at 16:53, Mark Newton newton@internode.com.au wrote:
On 22/08/2011, at 7:52 PM, RAKESH MOHAN AGARWAL wrote:
In this process all economies will be taken care of and will not be left out like in case of IPv4 era,If your policy proposal had been adopted in, say, 1996, in what ways would the resulting allocation of IPv4 resources differ from what we have now?
- mark
-- Mark Newton Email: newton@internode.com.au (W) Network Engineer Email: newton@atdot.dotat.org (H) Internode Pty Ltd Desk: +61-8-82282999 "Network Man" - Anagram of "Mark Newton" Mobile: +61-416-202-223
sig-policy: APNIC SIG on resource management policy *
sig-policy mailing list sig-policy@lists.apnic.net http://mailman.apnic.net/mailman/listinfo/sig-policy
Hi,
On 22 August 2011 21:23, Mark Newton newton@internode.com.au wrote:
On 22/08/2011, at 7:52 PM, RAKESH MOHAN AGARWAL wrote:
In this process all economies will be taken care of andwill not be left out like in case of IPv4 era,
If your policy proposal had been adopted in, say, 1996, in what ways would the resulting allocation of IPv4 resources differ from what we have now?
My prediction:
If in 1996, the IPv4 address pool had been carved up into 100+ per-country allocations (whether based on population, or GDP, or annual technology spend ...) then, long before 2011, some countries' address pool would have run dry, while many others would still have lots left. The total of the pockets of addresses in each pool would be a very large amount (say 50% of total). But these addresses would be unavailable to people (in other countries) who really needed them.
It would have been impossible to correctly predict in 1996 which countries would have, for example, the biggest uptake of smartphones in 2011 and allocate pool sizes correctly back then.
John
Hi,
On 22 August 2011 20:22, RAKESH MOHAN AGARWAL ddgnt-dot@nic.in wrote:
Dear All****
I have seen lot of comments and debate for Prop.100 and from some ofthe comments , it is clear some of the members are not clear about the very purpose of the proposal , so it is requested to all to kindly go through our original proposal and clarifications issued on 17/8/2011 then it will be clear to them that this proposal is not talking only of some specific countries, but taking care of all economies in the APNIC Region, as we have said in our original proposal that country wise Address block can be allocated to NIRs( For those economies those who are having NIR) and further allocation to Different organizations can be done by the NIR and of course through APNIC, keeping alignment with APNIC policy and for non NIR countries this big block can be reserved in APNIC secretariat and allocations to different organizations in that country can be done out of this big block.
I think you and I have fundamentally different ideas about how the Internet works. It is not "economies" that apply / pay for / use IP addresses; it is organisations. It is not "economies" that advertise routes into the global routing table, make routing policy decisions, be dual homed etc; it is organisations.
The Internet is not the telephone network where each countries' government created/controls a monopoly telco, the phone number plan has a leading country prefix code, and peering happens country-to-country etc etc
[ See my previous message about why advertising a single whole of country IP address block won't be acceptable to users. I'll add one more reason -- bad things can happen when IPv4 and IPv6 routing aren't approximately the same. ]
****
In this process all economies will be taken care of andwill not be left out like in case of IPv4 era, while maintaining the control of RIRs ,This was the theme of the whole proposal . Many of the reputed members started discussions on the technicalities of the advantages, as quoted in the proposal and opposed the proposal on these grounds , It is suggested to all APNIC community members to see the theme and spirit of the proposal as explained above and work towards overall interest of all countries and organizations in APNIC region.
Two years ago, there may have been 300 million free IPv4 addresses, and the gap between India's (current or predicted) data users and IPv4 addresses may have been about 300 million. Why didn't Indian ISPs apply for the some/most of the 300 million addresses?
Did you see my earlier question?
Q: Why didn't Indian ISPs get more IPv4 address space? a) they didn't know IPv4 addresses were going to run out so fast? b) they asked for it, but weren't granted it (insert conspiracy theory here)? c) they wanted to ask for it, but it was too expensive (10c per IP per year)? d) they planned to provide network service using NAT/proxies? e) other _____
Have there been any reports issued analysing why India didn't get lots of IPv4 addresses? Any academic papers, conference presentations, or Internet Drafts?
Once we know why the problem occurred, then we can look at how to prevent it happening again.
For example, if the problem was (a) -- then better education/planning is the solution (b) -- then maybe the allocation policy/procedure needs changing (c) -- much less important now since IPv6 addresses are much cheaper than IPv4 addresses (d) -- maybe the ISPs will want NAT for IPv6 too, and don't need IPv6 addresses !! (e) -- some other solution
IPv6 addresses are at least a billion times more abundant than IPv4 addresses. Maybe after they are 25% used up, then we could think about reserving some e.g. for ISPs that haven't been created yet.
Until then, please explain what is broken about the current policy: a organisation creates a network plan, applies for IPv6 addresses based on need (or uses the 1-click apply for a /32), pays their annual membership+usage fee to APNIC
This scheme treats all applicants equally. Assignment based on need. No discrimination against or for applicants from any country.
Until we get some agreement about the reasons why various countries didn't get enough IPv4 addresses, and what problems will stop them getting enough IPv6 addresses, we won't make any progress on agreeing on how to solve those problem.
John
On Aug 22, 2011, at 9:00 AM, John Mann wrote:
Hi,
On 22 August 2011 20:22, RAKESH MOHAN AGARWAL ddgnt-dot@nic.in wrote: Dear All
I have seen lot of comments and debate for Prop.100 and from some of the comments , it is clear some of the members are not clear about the very purpose of the proposal , so it is requested to all to kindly go through our original proposal and clarifications issued on 17/8/2011 then it will be clear to them that this proposal is not talking only of some specific countries, but taking care of all economies in the APNIC Region, as we have said in our original proposal that country wise Address block can be allocated to NIRs( For those economies those who are having NIR) and further allocation to Different organizations can be done by the NIR and of course through APNIC, keeping alignment with APNIC policy and for non NIR countries this big block can be reserved in APNIC secretariat and allocations to different organizations in that country can be done out of this big block.I think you and I have fundamentally different ideas about how the Internet works. It is not "economies" that apply / pay for / use IP addresses; it is organisations. It is not "economies" that advertise routes into the global routing table, make routing policy decisions, be dual homed etc; it is organisations.
The Internet is not the telephone network where each countries' government created/controls a monopoly telco, the phone number plan has a leading country prefix code, and peering happens country-to-country etc etc
[ See my previous message about why advertising a single whole of country IP address block won't be acceptable to users. I'll add one more reason -- bad things can happen when IPv4 and IPv6 routing aren't approximately the same. ]
In this process all economies will be taken care of and will not be left out like in case of IPv4 era, while maintaining the control of RIRs ,This was the theme of the whole proposal . Many of the reputed members started discussions on the technicalities of the advantages, as quoted in the proposal and opposed the proposal on these grounds , It is suggested to all APNIC community members to see the theme and spirit of the proposal as explained above and work towards overall interest of all countries and organizations in APNIC region.Two years ago, there may have been 300 million free IPv4 addresses, and the gap between India's (current or predicted) data users and IPv4 addresses may have been about 300 million. Why didn't Indian ISPs apply for the some/most of the 300 million addresses?
Did you see my earlier question?
Q: Why didn't Indian ISPs get more IPv4 address space? a) they didn't know IPv4 addresses were going to run out so fast? b) they asked for it, but weren't granted it (insert conspiracy theory here)? c) they wanted to ask for it, but it was too expensive (10c per IP per year)? d) they planned to provide network service using NAT/proxies? e) other _____
Have there been any reports issued analysing why India didn't get lots of IPv4 addresses? Any academic papers, conference presentations, or Internet Drafts?
Once we know why the problem occurred, then we can look at how to prevent it happening again.
For example, if the problem was (a) -- then better education/planning is the solution (b) -- then maybe the allocation policy/procedure needs changing (c) -- much less important now since IPv6 addresses are much cheaper than IPv4 addresses (d) -- maybe the ISPs will want NAT for IPv6 too, and don't need IPv6 addresses !! (e) -- some other solution
IPv6 addresses are at least a billion times more abundant than IPv4 addresses. Maybe after they are 25% used up, then we could think about reserving some e.g. for ISPs that haven't been created yet.
Until then, please explain what is broken about the current policy: a organisation creates a network plan, applies for IPv6 addresses based on need (or uses the 1-click apply for a /32), pays their annual membership+usage fee to APNIC
This scheme treats all applicants equally. Assignment based on need. No discrimination against or for applicants from any country.
Until we get some agreement about the reasons why various countries didn't get enough IPv4 addresses, and what problems will stop them getting enough IPv6 addresses, we won't make any progress on agreeing on how to solve those problem.
Hi John,
Not exactly an academic research paper, but then the empirical/historical facts are not really in dispute:
http://www.circleid.com/posts/ip_address_allocation_vs_internet_production_i...
Note that this was written back in 2005, and was intended to serve as a crude "how-to" primer for aggrieved/misinformed parties, as much it was a critique of the misunderstanding itself. For several years it was the most "widely viewed" article on CircleID, although I think it lost that distinction some time ago.
FWIW, I basically agree with your earlier "prediction," but I think it could be stated even more strongly:
Begin forwarded message:
My prediction:
If in 1996, the IPv4 address pool had been carved up into 100+ per-country allocations (whether based on population, or GDP, or annual technology spend ...) then, long before 2011, some countries' address pool would have run dry, while many others would still have lots left. The total of the pockets of addresses in each pool would be a very large amount (say 50% of total). But these addresses would be unavailable to people (in other countries) who really needed them.
It would have been impossible to correctly predict in 1996 which countries would have, for example, the biggest uptake of smartphones in 2011 and allocate pool sizes correctly back then.
If the IPv4 address pool had been allocated along national lines as described, then it seems equally plausible that recipient countries would still be holding an abundance of IPv4 addresses -- and not just today, but forevermore -- as the early, relatively geographically-concentrated Internet production/expansion booms that triggered rising global interest in the Internet participation would have never happened, or would have been dampened much sooner by the (artificial) scarcity of useful Internet addresses. Had that happened, perhaps all countries might have been spared the inconvenience and indignity of lacking the (x) hundred million IPv4 addresses required to support ubiquitous smartphone deployment -- because there would be no equivalent global e2e-style network that would have created a corresponding requirement for phones to be "smart" in the first place.
Regards,
TV
On Monday, August 22, 2011 02:32:16 PM Naresh Ajwani wrote:
More so, next hop address tracking is in the transition stage and not the ecosystem. The prop 100 is for NIRs and shud be debated in accordance only.
If a network is running v6 seriously, chances are they're using an improved next-hop tracking implementation on a Cisco-based platform.
However, I agree that this should not be used as an argument to promote Prop-100, as we'd be speaking on behalf of non- Cisco/newer-IOS users that would otherwise not support this proposal. The faults of some of our vendors, in this particular case, should not necessarily guide policy.
Mark.
On Monday, August 22, 2011 01:45:29 PM ramesh.chandra@airtel.in wrote:
Hi Mark, agree, there is lot of improvement in SW coding. Cisco has also improved in XOR release. However, number of routers running XOR is very low. Cisco as market leader runs same process in IOS used in most of routers at present. Same is with other vendors and impact possibly may be different. There is no intention of summarizing routers available in different AS.
The improvements are I speak about are not unique to IOS XR, but to IOS and IOS XE as well.
That said, the BGP Scanner is a feature unique to Cisco. Other vendors such as don't necessarily implement next-hop tracking the same way.
However, it is conceivable that any network running an IOS- based router that's implementing IPv6 seriously today is supporting a recent IOS release that has done away with the old BGP Scanner. This likelihood is significantly increased if an ISP plans to support 4-byte ASN's natively, but I digress.
While the next-hop tracking discussion is slightly academic owing to the fact that it's not possible to know what code the majority of the Internet is running, I don't think it's reasonable to use that as an argument to promote this proposal either.
Mark.
Activity Summary
- 4307 days inactive
- 4307 days old
- sig-policy@lists.apnic.net
- 7 participants
- 9 comments