Re: [IETFMIBS] Extending InetAddress(Type) for BGP Multicast VPNs

> I am not sure what would be
> meant by
> "two interoperating independent implementations" of a MIB module).

Roughly - the method used by the WGs developing MIB modules was to test
(at least) two independent implementations of agents and two
implementations of manager applications, run (at least) a complete MIB
walk on the respective MIB module and make sure that the results
obtained by the two applications from the two agents are consistent. 

Regards,

Dan

> -----Original Message-----
> From: ietfmibs-bounces@ietf.org [mailto:ietfmibs-bounces@ietf.org] On
> Behalf Of Adrian Farrel
> Sent: Tuesday, January 17, 2012 8:49 PM
> To: 'Jeff Haas'; 'Joan Cucchiara'
> Cc: ietfmibs@ietf.org
> Subject: Re: [IETFMIBS] Extending InetAddress(Type) for BGP Multicast
> VPNs
> 
> Hi,
> 
> The "two implementations" rule is imposed on a case-by-case basis by
> the IDR WG
> chairs. It is usual for protocol modifications to BFD.  I am not aware
> of it
> being enforced for MIB modules (indeed, I am not sure what would be
> meant by
> "two interoperating independent implementations" of a MIB module).
> 
> Let's allow the WG chairs to decide what is needed in this case.
> 
> Cheers,
> Adrian
> 
> > -----Original Message-----
> > From: ietfmibs-bounces@ietf.org [mailto:ietfmibs-bounces@ietf.org]
On
> Behalf
> > Of Jeff Haas
> > Sent: 17 January 2012 15:46
> > To: Joan Cucchiara
> > Cc: ietfmibs@ietf.org
> > Subject: Re: [IETFMIBS] Extending InetAddress(Type) for BGP
Multicast
> VPNs
> >
> >
> > On Jan 17, 2012, at 10:02 AM, Joan Cucchiara wrote:
> > > Could you remind me again what the requirements are by the the IDR
> WG to
> > > have
> > > a document advance to RFC status?   As I recall, this is more
> involved than
> > > what is required by
> > > other WGs.
> >
> > Two independent implementations.
> >
> > > Could you also tell us what the status is of the BGPv2 MIBs with
> respect to
> > > the IDR WG requirements?
> >
> > Stable, but lacking implementation of the latest draft.  Juniper has
> an
> enterprise
> > MIB implementation of the older draft but I haven't had the
> opportunity to
> > update the implementation since joining.  (MIBs get poor attention
> from
> product
> > managers in general.)  Another large vendor is rumored to have an
> > implementation in progress.
> >
> > I also had most of an implementation in Quagga that I have yet to
> complete.
> The
> > majority of my problem was net-snmp issues with notifications,
> particularly
> with
> > sysUpTime timestamps.
> > I probably should complete that just to push the standard.
> >
> > -- Jeff
> >
> > _______________________________________________
> > IETFMIBS mailing list
> > IETFMIBS@ietf.org
> > https://www.ietf.org/mailman/listinfo/ietfmibs
> 
> _______________________________________________
> IETFMIBS mailing list
> IETFMIBS@ietf.org
> https://www.ietf.org/mailman/listinfo/ietfmibs