<<< Chronological >>> Author Index    Subject Index <<< Threads >>>

Re: IP Management Tool - Minimum Requirements

  • To: Michael van Elst < >
  • From: Dmitry Morozovsky < >
  • Date: Fri, 20 Apr 2001 19:35:41 +0400 (MSD)
  • Cc: Guy Vegoda < >

On Thu, 19 Apr 2001, Michael van Elst wrote:

MvE> The tool will accept a number for the count of IP addresses to be assigned
MvE> and search for a suitably aligned assignment. Obviously you will end with
MvE> holes in your allocation whenever an assignment is returned.
MvE> 
MvE> There are various alogrithms (coming by memory or disk space allocators)
MvE> that try to get 'optimal' results by filling the holes. This tool does a
MvE> two phase search, first it looks for assignments that fill holes completely
MvE> and then it falls back and searches the first hole in the allocation that
MvE> is large enough.
MvE> 
MvE> 
MvE> > I don't want an end product that decides for you
MvE> > which subnet you will be assigning.
MvE> 
MvE> Why not ? Most assignments are completely arbitrary. There are two
MvE> reasons to chose a specific subnet: aggregation and reservation.
MvE> In both cases you do a bigger assignment first (which only exists
MvE> privately, not in the RIPE database) and select your assignments
MvE> from this block.

I agreed, and we tried to first fill requests for, e.g., /24 in even
blocks in /20 "reserved" for /24s, to be expanded to /23 when needed. Same
algorithms for other popular block sizes, such as /27 and /26...


Sincerely,
D.Marck                                   [DM5020, DM268-RIPE, DM3-RIPN]
------------------------------------------------------------------------
*** Dmitry Morozovsky --- D.Marck --- Wild Woozle --- marck@localhost ***
------------------------------------------------------------------------





  • Post To The List:
<<< Chronological >>> Author    Subject <<< Threads >>>