[RC5] Where do the packets disappear!??

Steve Lee rc5 at banoffee.demon.co.uk
Thu Sep 24 19:59:47 EDT 1998


In message <19980921113608.H25625 at jeep.ops.neosoft.com>
          Daniel Baker <dbaker at jeep.ops.neosoft.com> wrote:

> On 21-Sep-1998, WINKELMANS Gunther wrote:
> > A few months ago the keyserver was not big enough to hold all blocks. 
> > Instead it could only contain 4 subspaces.
> 
> The keymaster has been greatly expanded thanks to user donations.  We
> currently have 12 keyspaces mounted and have the capability of having
> up to 40.
> 
> > This resulted in blocks being recycled after ONE MONTH: subspace no. 5 
> > could only be started when subspace no. 1 was completely terminated.
> > 
> > At that time everybody was advised to throw inbuffers away after one month 
> > and to get fresh ones.
> 
> It has been upgraded and keyspaces are left to sit for at least six
> months before they are recycled.

ISTM that such a strategy would also be appropriate for randomly allocated
blocks.

Before the keyservers changed from dealing out 0x71 blocks to dealing
out 0x72 blocks, any machines which needed to generate random blocks
were being directed to the 0x72 subspace.

As the 0x72 keyspace was handed out immediately afterwards, there would
almost certainly be duplication between random and allocated blocks.

Surely it would not be too difficult to ensure that there is a suitable
delay between clients being directed towards a subspace for generating
random blocks and the subspace being used for allocated blocks?

-- 
Steve Lee
http://www.banoffee.demon.co.uk/steve/

--
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest



More information about the rc5 mailing list