Re: nhrp-05 - Purge packets

[Andrew Smith <[e-mail]>]

> From [e-mail] Fri Nov  3 09:08:43 1995
> To: [e-mail] (Andrew Smith)
> Subject: Re: nhrp-05 - Purge packets 
> Cc: [e-mail]
> Date: Fri, 03 Nov 1995 11:51:38 -0500
> From: James Luciani <[e-mail]>

Jim,

> > Doesn't the receiver also have to match on some other things e.g.
> 
> I am strongly against keeping this much state around on a per cache entry
> basis.  I disagee. (28 bytes for source info X number of cache entries).

Sorry, but the protocol has to work correctly. Life's tough. If you can show
that it works correctly without these checks under conditions where the NHS is 
failing over then that's fine. I don't want to see another 1/2-engineered
protocol out there: we have enough of those around here already.

...
> Don't we know the NBMA Subnet ID as a result of knowing which interface
> the packet came in on?  A wrong NBMA Subnet ID is in fact an error condition.
> > - what about NBMA Subnetwork ID? Nowhere is it mentioned that this
> >   is or is not allowed in a Purge: presumably it does need to be
> >   there in the Purge if an implementation is using it in Request/Reply 
> >   packets. If it's there, it needs to be checked.

So you're saying that it's already checked on reception? That wasn't clear
to me from reading the current spec.

> 
> Regards,
> -- Jim Luciani

Andrew


********************************************************************************
Andrew Smith					TEL:	[phone]
Technology Synergy 				FAX:	[phone]
Bay Networks, Inc.				E-m:	[e-mail]
Santa Clara, CA
*********************************************************************************