[ncc-services-wg] Certifying of PI End User Address Space

On 25 Mar 2013, at 10:55, Andrew de la Haye wrote:

> For option 2 (going through the sponsoring LIR) and option 3 (going to the RIPE NCC directly), there is a possibility to provide an automated solution for which only PI End Users who have submitted a 2007-01 End User Assignment Agreement are eligible. This means the RIPE NCC will need to develop a solution where the following informational elements are cross-referenced with each other:
> 
> a) The authoritative control over the address space (i.e. the ability to authenticate against the relevant objects in the RIPE Database)
> b) The End User Assignment Agreement that was submitted and verified by the RIPE NCC
> c) The RIPE NCC Access credentials for accessing the certification management interface
> 
> Please keep in mind that in both estimates, where the actual cost falls in the given broadband depends on the amount of manual verification that is desired, which can range from random periodic audits to verification of every application by RIPE NCC staff. Also, the additional costs with smaller or larger uptake are estimated to be quite low. 
> 
> Option 2: 
> A one time cost of 65 kEUR, with a running cost of 45 kEUR, based on an estimated uptake of 15% or approximately 2,500 PI End Users. 
> 
> Option 3: 
> A one time cost of 70kEUR, with a running cost of 70 kEUR, based on an estimated uptake of 15% or approximately 2,500 PI End Users.
> 
> Lastly, it is important to point out that as the certificate is valid for only one year, providing this service to PI End Users creates an opportunity for periodic confirmation of the 2007-01 End User Assignment Agreement, further improving the robustness of our Registry.

	Thank you for this useful information, Andrew.

	I'ld appreciate some further clarification, if you wouldn't mind.

	It seems to me that there must be significant overlap in the elements needed to implement either
	Option 2 or Option 3.  Should we understand that the estimates you give correspond to the respective
	costs of implementing either option on its own, without implementing the other?  If so, would it
	be reasonable to assume that implementing both might be done for a one-time cost nearer to €75k
	than to €135k and a running cost somewhere between your estimates, depending on the mix of
	PI End Users deciding for each of the options?

	Thanks in anticipation and best regards,
	Niall O'Reilly