Re: [address-policy-wg] 200 customer requirements for IPv6
To all the nay-sayers and ignorants in the die-hard IPv6 camp I can only recommend that they put yourself into the shoes of a large coporation with 10,000 employees for a day.
How many 10K employee organizations exist in the world today (or in ten years)? How many routing prefixes would be required in the DFZ to support this? Can the routing infrastructure handle that? If there were consensus that the answer to the above questions was "yes, it's manageable", I'm sure folk would support a modified policy to that effect in a heartbeat. But in the absence of some real data showing the implication of loosening the policy to allow the above, many (myself include) will be very concerned about the consequences of doing this. Thomas
How many 10K employee organizations exist in the world today (or in ten years)?
In the USA there are 930 such companies. This does not include non-commercial organizations. If you count companies with more than 2,500 employeees, then there are 3,316 in the USA. In the EU there are 10,000 large enterprises, that is companies and non-profits engaged in economic activity with more than 250 employees. I think that 10,000 employees is much too high as a cutoff for multihoming demand. Even 2,500 is too high. So let's say that there are 5,000 such enterprises in the USA and another 5,000 in Europe. This seems a reasonable guess. If we assume that the combined population of the EU and the USA is one fifth of the world's enterprise population, then we can expect five times 10,000 multihomers or 50,000 enterprises worldwide to multihome. On the face of it, that doesn't seem too bad but it misses some things. Government organizations. Smaller companies whose desire for multihoming does not come from the size of their business but from the nature of their business. For instance the family website that is running a transaction-based business that makes very slim margins on large numbers of transactions and therefore needs to be up 24x7. Or a web shop that has no storefront and relies on a steady stream of Internet orders, maybe a small taxi firm? This is a tough question to answer because it is entirely non-technical and we do not have the support from economists statisticians, and geographers to attempt a serious answer. If RIPE and the NRO would fund some studies, we would all be a lot better off in deciding the right policies.
If there were consensus that the answer to the above questions was "yes, it's manageable", I'm sure folk would support a modified policy to that effect in a heartbeat.
But in the absence of some real data showing the implication of loosening the policy to allow the above, many (myself include) will be very concerned about the consequences of doing this.
The wording could be changed without loosening it up much if we note that "customers" could refer to internal business units, not just external customers. --Michael Dillon
participants (2)
-
Michael.Dillonļ¼ btradianz.com
-
Thomas Narten