[governance] PP: India wants to abolish BGP and introduce national routing and IP management

Mon Oct 27 01:22:52 EDT 2014

Below tagged [srs].   And speaking only for myself and not for any organization whatsoever.

From: governance-request at lists.igcaucus.org [mailto:governance-request at lists.igcaucus.org] On Behalf Of Guru Acharya
Sent: Monday, October 27, 2014 9:52 AM
To: governance at lists.igcaucus.org; Salanieta T. Tamanikaiwaimaro
Cc: michael gurstein; JNC Forum
Subject: Re: [governance] PP: India wants to abolish BGP and introduce national routing and IP management

I agree that India's strategy in Proposal 98 is not well thought out. However, I think India's concerns stem from the following:

1)      IP addresses are not equitably distributed in the Asia Pacific region. The skewed allocation is reflected in the statistics that Eastern Asia holds 2,712,098 of the IPv4/24 addresses while South Asia (including India) holds only 170,365 of the IPv4/24 addresses.

[srs]  IP addresses are allocated according to need.   And you forget that East Asia includes, for example, China, which has a surprisingly large amount of IP address space – and of course, they also have excellent broadband penetration.  There are villages in South China that may have one bus a day to the big city but even those have broadband.   Compare that to a much smaller country in South Asia like say Laos or Sri Lanka.

This is not a “one country one fixed ration of IP addresses” setup – that might work for India’s public distribution system, but not much else.

2) APNIC Executive Council (EC) has remained largely static and arguably captured by the East Asians and Australians for almost a decade. In the APNIC EC elections, the votes allotted to members are in proportion of the IP addresses held by them. For example, if the IP holding is up to /22, the member has 2 votes; and if the IP holding is between /13 and /10, then the member has 32 votes.  Effectively, due to the current skewed allocation of IP addresses, representatives of India do not stand much of a chance in APNIC EC elections. Notably, India has had just one representative (for one year) on the APNIC EC in the past decade. Additionally, while this system of proportional voting creates a bias in favour of incumbent members who have grandfathered large IP holdings, the system penalises those members who are using IP addresses efficiently (for example by using Network Address Translation) and also penalises the community that is yet to connect to the Internet or has connected to the Internet late. 

[srs] Having been to several APNIC meetings before, I was struck by the fact that so few Indian ISPs even bothered to travel to APNIC whereas I would find Vietnamese, Laotians, Sri Lankans, Pakistanis [etc] .. and even where Indians have stood for the EC before, they have often  been “non technical CEO” type candidates.    Possibly a different kind of person attending APNIC and gaining recognition by participating in the various SIGs might have a great chance for an EC role.  I was myself on the board of APRICOT for several years while I was able to participate in it.   

Before you say India is from the global south and travel options are few and limited, let me also point you to the fact that there are plenty of people who apply for APNIC / APRICOT / SANOG fellowships from countries like Nepal, Sri Lanka and Bangladesh, I might see three or four Indian fellowships, if at all.   Mostly, in proportion to the number of Indian attendees at these events.   

3) There are two options for redistribution of IP addresses. The first is to go through the APNIC PDP, which is to reform APNIC from within. The second is to bypass APNIC and ask ITU to take over the RIR function. India seems to have adopted the second path due to lack of trust in the first path, which would be slow, bottom-up, and subject to resistance by incumbents. Further, APNIC EC plays a crucial role in the consensus building process and I doubt any reform of the APNIC EC will not be resisted.

[srs] India already accepted the APNIC process to the extent of [1] boosting participation by Indian ISPs in APNIC [2] forming a NIR (National Internet Registry) currently administered by NIXI etc.    Also – how many APNIC meetings have you attended, or is this like our other “friend” from India who keeps criticizing ICANN without having attended a single ICANN meeting?

4) With respect to the IANA transition, the APNIC secretariat drafted a proposal and pushed that proposal top-down onto the community, which was accepted as having consensus without any intelligent discussion in a conference (APNIC38) at a remote location (please read transcripts of APNIC38). This proposal suggests NTIA oversight should be replaced with a SLA/AOC between ICANN and the NRO (combination of the 5 RIRs). An obvious corollary of this extra added responsibility of oversight should be enhanced accountability of the RIRs. Notably, APNIC is refusing to accept any discussions on enhancing its accountability as part of the IANA transition plan. Enhanced accountability of APNIC would include a measure of representativeness in my opinion.

[srs] APNIC 38 was held last month in Brisbane.   I assume people from the pacific islands, say, are not entitled to have a short trip to an APNIC meeting at least once?   India is not more than a 4 or 5 hour flight from most other APNIC locations.

While I feel that India's concerns are genuine, I also feel the path adopted is incorrect. If this proposal goes through as is, it can fragment the Internet through three routes: First, through alternate (non-IETF) standards emerging (from ITU) to address security concerns that are not inter-operable with existing standards; Second, through a broken non-unique allocation of IP addresses where ITU and RIRs allocate IP addresses in parallel; Third, through an alternate root zone emerging to address the names part of Proposal 98.

[srs] The technical implementation remains to be seen.  However, along with standards, may I also add that appropriate hardware / software (routing, tcp/ip stacks etc) that support these new standards will also have to be developed, especially if there is significant divergence from existing global standards.

Additionally, some of India's concerns are driven by cyber security that can be addressed domestically and do not require regional/international reform. For example, domestic routing of traffic can be done by making the IXPs more efficient. The comments received in the TRAI (India's telecom regulator) consultations point towards numerous areas of improvement for NIXI (India's IXP). Similarly, domestic hosting of content can be facilitated by improving the privacy regime and intermediary liability regime.

[srs] This is actually part of the problem.   The other part is trust and competition issues among Indian ISPs that are also historically significant interconnection barriers. 

--srs
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.igcaucus.org/pipermail/governance/attachments/20141027/b86a5eb2/attachment.htm>
-------------- next part --------------
____________________________________________________________
You received this message as a subscriber on the list:
     governance at lists.igcaucus.org
To be removed from the list, visit:
     http://www.igcaucus.org/unsubscribing

For all other list information and functions, see:
     http://lists.igcaucus.org/info/governance
To edit your profile and to find the IGC's charter, see:
     http://www.igcaucus.org/

Translate this email: http://translate.google.com/translate_t