[ppml] Definition of "Existing Known ISP"
Owen DeLong
owen at delong.com
Thu May 3 05:20:38 EDT 2007
- Previous message: [ppml] Definition of "Existing Known ISP"
- Next message: [ppml] Definition of "Existing Known ISP"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On May 3, 2007, at 1:38 AM, <michael.dillon at bt.com> <michael.dillon at bt.com> wrote: >> In general, if you are assigning /32s to customer utilization, but, > not >> to independently routed segments (e.g. you have a bunch of customers >> on the same server each assigned a /32), then, no, you don't meet >> ARIN's definition of an ISP. > > This is news to me. > > Since when does ARIN tell ISPs what their business model must be? I > thought that ARIN's policies required an org to give "technical > justification" for IP address assignments and that SSL webhosting > counts > as a technical justification. Where does the policy prohibit assigning > addresses to customers on the same server? Michael, It's got nothing to do with business models. It has to do with network topology. From an ARIN perspective, reassignments apply to networks, not to host addresses. If you're reassigning network segments to other entities, then, you're in the ISP definition. If you're not, then, you're probably in the end-user definition. This has nothing to do with what is or isn't legitimate address usage. It has to do with a definition of which section of the policy you fall under. The SSL webhosting counts as technical justification for end user networks where host counts and technical justification for efficient utilization within each given network are what matters. In the ISP section, what matters is the portion of network space you have assigned/reassigned and that all reassignments and assignments meet the end-user guidelines for efficient utilization. In short, ISPs are responsible for conforming to two sets of guidelines and making sure that their customers conform to one of those two sets. End users only have to conform to the one set. This is an absurd semantic interpretation of my statement and I don't understand what you think is gained for anyone by muddying the waters with it. I have already agreed that the usage of the term ISP in the policy to represent this concept is a flawed use of the term. However, absent a policy proposal to correct that wording (which I believe will probably come up shortly from another source), this thread was started as an attempt to clarify the meaning of the term FOR PURPOSES OF THE EXISTING ARIN POLICY. It's pretty clear from the context and historical application of ARIN policy that: 1. IPv4 resources issued by ARIN fall into two categories. A. Organizations who issue networks or blocks of networks to other organizations (reassignments and/or reallocations). B. Organizations who directly consume network addresses on machines owned/operated by the organization, even if those addresses are unique to particular customers of the organization. 2. The primary actual implications of the distinction between these policies fall into two categories. A. Membership -- Organizations in category 1A receive ARIN membership as part of their resource subscription fees. Organizations in category 1B do not. They can purchase ARIN membership as a separate fee. B. Fees -- Organizations in category 1A pay generally higher fees based on the total amount of resources held by the organization. They do not pay initial allocation fees, but, they do pay any applicable increase in their renewal fees upon issuance of any resource which moves them into a different fee category. Organizations in category 1B pay a flat fee of $100/year for maintenance of all of their resources, but, each request for additional resources results in an initial assignment fee for said resource(s). 3. This distinction is more about how the addresses are treated topologically than it is about business models, the actual definition of an ISP in the real world, the phase of the moon or the price of tea in China. Now, the question at hand refers to the section of ARIN policy regarding initial allocations of IPv6 address space. I believe that the clear intent of the IPv6 policy as written was to make it a no-brainer for ARIN to issue /32 allocations to any organization known to ARIN as an IPv4 reassigner/reallocator. Owen -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 2105 bytes Desc: not available Url : http://lists.arin.net/pipermail/arin-ppml/attachments/20070503/75d11760/attachment.bin
- Previous message: [ppml] Definition of "Existing Known ISP"
- Next message: [ppml] Definition of "Existing Known ISP"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the PPML mailing list