lowering maximum assignment window

On 11-Feb-99 Paul Tate wrote:
>  
>   Stephen Burley <stephenb at uk.uu.net> writes:
>   * Can I make a suggestion: 
>   * 
>   * We leave the maximum as a /19 and for a period of time lower the
>  registries
>   * assignment window to a reasonable level i.e. a /22 or /23 which will give
>  ea
>   * ch
>   * registry the time to prove their procedures and expose them to RIPE
>  policies
>   * again. This could be done on a rotational basis so each registry could
>  prove
>   *  to
>   * RIPE that they have good working practice. Once RIPE was happy with the
>   * registry they would raise the assignment window to the original size.
>  This
>   * could be done on a 2 year basis. So within a 2 year period RIPE would be
>  hap
>   * py
>   * that any given registry is applying RIPE policy and DB correctly. Of
>  course
>   * this would not need to be done to registries who RIPE are already in
>  regular
>   * contact with.
>  
>  I think the above comments defeat the purpose of the suggestion to cap the 
>  size of a maximum assignment window to, say a /21. I cannot imagine the 
>  increase in workload on registries like yourself who have a /19 aw to be all
>  that great. Customers requesting a /20 and /19 worth of address space must
>  be 
>  in a minority and besides a turnaround time of three days for such requests
>  is 
>  pretty reasonable.
Not really. This not an academic arena, people are paying for the best service
they can get and want their connection tomorrow i am not joking. This reminds
me of the UK name space requests which had a 3 to 5 day turn around which was
unacceptable so nominet was created and took over the UK name space, we now
have instant registration. Just an example of why 3 days is not an acceptable
delay in installation time i think your comment was a little naive. 

 Your comments about monitoring registries on a two year 
>  basis would, I feel, seriously increase the burden on us here at RIPE NCC. I
>  think you should also bear in mind that the suggestion to lower the maximum
>  aw 
>  is also an attempt to make the playing field more even so to say.
>  

I disagree, the lowering of the largers registries aw to a lower size will
probably generate just as much work for ripe. Also if you only apply this to
the largest aw's then what about the registries with aw of less than /21, they
could be making as many mistakes and not applying policies, all registries must
be equaly treated. By rotating the lowering as in Paula's last email then all
registries are checked for policy compliance. 


>  Paul Tate
>  Hostmaster
>  RIPE NCC
>  
>  

----------------------------------
Stephen Burley
Senior Hostmaster for UUNET

Date: 12-Feb-99
Time: 08:50:07

http://www.uk.uu.net
----------------------------------
An MCI WorldCom Company