This policy proposal has been accepted
The new RIPE Document is: ripe-655
The current RIPE IPv6 Address Allocation and Assignment Policy (ripe-641, section 5.1.2) recognises only “the number of existing users and the extent of the organisation’s infrastructure” when assessing the requirement for an initial allocation size greater than /29. This is proving problematic for organisations whose IPv6 addressing requirements are not fully represented by such criteria.
For example:
Hierarchical Networks – The infrastructure of some organisations, particularly those with a multi-national presence, is often made up of component networks which can be regarded as Internet service providers in their own right. In many cases these ISPs could be eligible to become LIRs themselves and therefore each could obtain a /29 without justification. However, it is often important that they operate subordinately to the ‘parent’ organisation not only administratively but also technically on matters such as routing and addressing. In such cases, the need for a hierarchical addressing strategy can lead to an overall addressing requirement larger than that measured simply by the number of End User networks yet the current assessment criteria does not allow consideration of this, despite the principles of hierarchy and aggregation being stated (in section 3) as key goals of address space management.
Multiple Discrete Networks – Some organisations have a requirement to operate multiple discrete networks with completely independent routing and address management policies. Justification varies but may include, for example, regulatory restrictions on data transmission or geographic distance/diversity between networks. However, the requirement for multiple discrete prefixes is not eligible for consideration by the current initial allocation size assessment criteria.
It is therefore proposed that the assessment criteria for an initial allocation size greater than /29 should allow consideration of aspects other than “organisation size” alone. To achieve this, it is proposed that the criterion “number of existing users and the extent of the organisation’s infrastructure” is expanded with other factors that can be considered in order to accommodate the justified needs of a broader range of organisations within the RIPE community.
[The following text will update section 5.1.2 in the RIPE Policy Document “IPv6 Address Allocation and Assignment Policy“ if the proposal reaches consensus.]
Organisations that meet the initial allocation criteria are eligible to receive an initial allocation of /32. For allocations up to /29 no additional documentation is necessary.
Organisations may qualify for an initial allocation greater than /29 by submitting documentation that reasonably justifies the request. If so, the allocation size will be based on the number of existing users and the extent of the organisation's infrastructure.
Organisations that meet the initial allocation criteria are eligible to receive an initial allocation of /32. For allocations up to /29 no additional documentation is necessary.
Organisations may qualify for an initial allocation greater than /29 by submitting documentation that reasonably justifies the request. If so, the allocation size will be based on the number of users, the extent of the organisation's infrastructure, the hierarchical and geographical structuring of the organisation, the segmentation of infrastructure for security and the planned longevity of the allocation.
The assessment criterion “number of existing users” is arguably unambiguous and serves as one reasonable measure of the “size” of an organisation. However, the term “extent of the organisation’s infrastructure” is somewhat ambiguous and therefore open to interpretation. Even when using a dictionary definition of the term “extent” to mean area, volume, scope, etc., it is clear that again it is very much a measurement of “size”.
For many organisations (ISPs in particular), there is often a straightforward correlation between their “size” and the amount of IPv6 address space they therefore require. However, some organisations have legitimate addressing requirements that arise from factors other than outright “size” alone and yet the current policy, as written, does not allow consideration of such requirements to be made. The proposed expansion of the assessment criteria is intended to accommodate the justified needs of a broader range of organisations within the RIPE community in a way that can be consistently and fairly applied by the RIPE NCC’s IP Resource Analysts.
Note: In order to provide additional information related to the proposal, details of an impact analysis carried out by the RIPE NCC are documented below. The projections presented in this analysis are based on existing data and should be viewed only as an indication of the possible impact that the policy might have if the proposal is accepted and implemented.
It is the RIPE NCC's understanding that the proposal would expand the assessment criteria used to evaluate IPv6 allocations larger than /29. In addition to the existing criteria (amount of users and extent of the organisation's infrastructure), the proposal introduces new criteria to be considered by the RIPE NCC when evaluating requests. These criteria are the hierarchical and geographical structure of the organisation, the segmentation of infrastructure for security, and the planned longevity of the allocation.
LIRs would need to provide comprehensive documentation that reasonably justifies their request.
The policy does not define any specific limits for these new criteria. Instead, the proposal asks the RIPE NCC to provide its understanding of what is reasonable.
Based on experience gathered through the evaluation of previous requests, the RIPE NCC suggests that it would apply the following understanding of "reasonable" for the different criteria. For a better overview, the RIPE NCC’s understanding of the existing criteria is also included.
The RIPE NCC asks the RIPE community to provide feedback on this understanding to make it part of the PDP and to guide the policy implementation should the proposal be accepted.
The aim of this suggested understanding of “reasonably justified” is to allow the creation of a well-aggregated IPv6 addressing plan while avoiding wasteful practices.
Amount of Users
The LIR shall provide comprehensive documentation that shows the amount of end sites and what subnet size will be provided per end site. The current policy defines an end site as an end user who has a business or legal relationship with the provider.
Extent of the Organisation’s Infrastructure
The LIR shall provide comprehensive documentation that shows the need for IPv6 for their network (POPs, etc.) in addition to the number of end sites.
Hierarchical and Geographical Structuring of the Organisation
The LIR shall provide a comprehensive network topology that describes the division of the network in hierarchical and geographical sections. The RIPE NCC will ask for additional documentation to justify why a less address consuming hierarchy or topology can not be implemented.
If this network topology is justified, the RIPE NCC will consider up to one extra bit per hierarchical level or geographical segment as reasonable, on top of the documented need for that part of the network. This allows for aggregated addressing of future assignments in this part of the network.
Segmentation of Infrastructure for Security
The LIR shall provide comprehensive documentation that explains the segmentation of the network for security reasons. The RIPE NCC will ask for additional documentation to justify why the security requirements cannot be achieved with a less address-consuming approach.
If the segmentation of the infrastructure for security requirements is justified, the RIPE NCC considers up to one bit extra per security segment reasonable.
Planned Longevity
If the LIR wishes to include growth expectations in the requested allocation size, it shall provide comprehensive documentation about the past growth of its services. The RIPE NCC will consider longevity reasonable for a similar timeframe for which past growth was documented.
LIRs can include multiple assessment criteria in their request. The RIPE NCC will evaluate each criterion and make sure that they are balanced against each other. For example, if the criteria of longevity is applied to the number of end sites, no extra bit can be added to this level in a hierarchical addressing plan. Each assignment will be taken into account only once towards the total count of needed IPv6 space for an organisation and will not be multiplied by the times it is encapsulated in higher addressing plan levels.
It is important to note that the suggested policy change could also have impact on the evaluation of subsequent IPv6 allocation requests. LIRs that request an initial allocation under the proposed new assessment criteria might find it difficult to reach a sufficient HD-ratio in the future, as a significant amount of address space will be reserved for network structuring and future growth. Meeting the specified HD-ratio is required to qualify for a subsequent IPv6 allocation.
Looking solely at the number of large IPv6 allocations requested in the past, the RIPE NCC would expect no significant impact compared to the overall amount of regular IPv6 allocations. However, the proposal would allow more LIRs to justify large IPv6 allocations. This could result in an increase in large requests and an overall increase in IPv6 consumption. The RIPE NCC is therefore unable to predict whether the proposal would have a significant impact on the IPv6 address pool.
The proposed policy would allow organisations better aggregation in their addressing plan. As future growth in one specific network segment does not require renumbering to achieve aggregation or alternatively the assignment of a separate address range, over the longer term a slowing effect on routing table growth is possible. The RIPE NCC has no historical data to quantify if this will have a significant impact on the size or stability of the Internet global routing table.
To ensure that large IPv6 requests are evaluated consistently, Registration Services has a team of IP Resource Analysts (IPRAs) who evaluate such requests together. The IPRAs will review the submitted documentation and might request additional information if required. The details of this documentation will vary with the amount of assessment criteria that need to be considered. Registration Services will ask for some of the following documents:
After the IPRAs have finished their evaluation, the request will follow the Internet number resource escalation process.
The evaluation of larger requests requires considerably more work than requests of a /29 or smaller. Still, the overall change in the workload will be minor, as Registration Services is already applying this evaluation process to large IPv6 requests. Also, the overall amount of such requests is estimated to be low, based on around ten requests for large allocations received in the past twelve months.
The RIPE NCC’s Training Services Department has been contacted in the past by organisations who plan to request large allocations with hierarchical, geographical or security-related IPv6 addressing plans and would like a more tailored course that covers this. If this proposal is accepted, an increase in these kinds of training activities is expected.
The proposed policy would widen the possibilities for IPv6 deployment, especially for large, non-ISP organisations. If the proposal is accepted, the IPv6 Program Manager will oversee the development of a document to help large organisations build a scalable IPv6 addressing plan that fits the requirements of the policy. The document would be written, reviewed and maintained with help from the RIPE community.
After analysing the data that is currently available, the RIPE NCC does not anticipate that any significant impact will be caused if this proposal is implemented.
After analysing the data that is currently available, the RIPE NCC does not anticipate that any significant impact will be caused if this proposal is implemented.
After analysing the data that is currently available, the RIPE NCC does not anticipate that any significant impact will be caused if this proposal is implemented.
The RIPE NCC estimates that the implementation of this proposal would have a low impact.
The RIPE NCC would mainly update existing processes and supporting documentation for IPv6 allocation requests to evaluate under the new criteria.