Re: draft-ietf-ospf-iana-00

[John Flick <john.flick at HP.COM>]

Actually, this information is true for ISO ASN.1, but not for SNMP.
SNMP's SMI is an "adapted subset" of ASN.1.  That subset is defined
in RFCs 2578-2580.

From RFC 2578, section 3.5:

   An OBJECT IDENTIFIER value is an ordered list of non-negative
   numbers.  For the SMIv2, each number in the list is referred to as a
   sub-identifier, there are at most 128 sub-identifiers in a value, and
   each sub-identifier has a maximum value of 232-1 (4294967295
   decimal).

I don't believe we have run into any problems with these limits in
SNMP :-) .

So, I don't believe that draft-ietf-ospf-iana draft has a problem here.

John


Tom Petch wrote:
> <inline>
> Tom Petch
> ----- Original Message -----
> From: "Kireeti Kompella" <kireeti at JUNIPER.NET>
> To: <OSPF at PEACH.EASE.LSOFT.COM>
> Sent: Tuesday, April 05, 2005 12:07 AM
> Subject: Re: draft-ietf-ospf-iana-00
>
>
>
> > On Mon, 4 Apr 2005, Tom Petch wrote:
> >
> >
> > > I am curious about  the choice of four octet for enterprise code in 2.5.
> > >
> > > When used in SNMP, these codes have no effective limit on length;
> >
> > I was under the impression that each number in an OID sequence was a
> > 32-bit number.  Is that not right?
>
>
> Not so.  BER.1 encodes the sub-identifiers of an OID (after the first two) in
> what I think of as CCITT encoding, using 7 bits per octet with the other bit as
> a flag to say this is (or is not) the last octet.  So a 7 bit sub-identifier
> encodes in one octet, 14 bit in two and so on.
>
> I have never seen any other limit imposed and - knowing ISO - I would not expect
> there to be one (apart from the overall limit on length of a TLV which is
> 2**1008 octet - this is ISO:-).
>
> <snip>