[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: INET-Addresses not used in MPLS MIB Modules
On Fri, 31 Jan 2003, Wijnen, Bert (Bert) wrote:
> For various cases, they cannot use the INET-ADDRESS-MIB
> definitions. They were using different mechanisms at
> different places, which is not good. So I suggested that
> if they can indeeed not use the INET-ADRESS-MIB TCs, that
> they define their own TCs that every MPLS mib module can
> re-use. It is actually more TE (Traffic Engineering)
> related than pure MPLS... but ok...
>
> Seems they too RFC2851 as an example, and maybe not 3291.
That's not good. We went through a long discussion on the
mibs@ops.ietf.org list why the "registered immediately before"
stuff was not a good idea (if I recall correctly, it came up
in the context of the review of the MALLOC-MIB, which would
have been seriously hobbled by that constraint).
> This is what they came up with. Any further comments?
>
> TeHopAddress ::= TEXTUAL-CONVENTION
> STATUS current
> DESCRIPTION
> "Denotes a generic Tunnel hop address.
>
> A TeHopAddress value is always interpreted within the
> context of an TeHopAddressType value. The
> TeHopAddressType object which defines the context must
> be registered immediately before the object which uses
> the TeHopAddress textual convention. In other words,
> the object identifiers for the TeHopAddressType object
> and the TeHopAddress object MUST have the same length
> and the last sub-identifier of the TeHopAddressType
> object MUST be 1 less than the last sub-identifier of
> the TeHopAddress object."
> SYNTAX OCTET STRING (SIZE (0..16))
>
> TeHopAddressAS2 ::= TEXTUAL-CONVENTION
> DISPLAY-HINT "d"
> STATUS current
> DESCRIPTION
> "Represents a two octet AS number."
> SYNTAX Integer32 (0..65535)
>
> TeHopAddressAS4 ::= TEXTUAL-CONVENTION
> DISPLAY-HINT "d"
> STATUS current
> DESCRIPTION
> "Represents a four octet AS number."
> SYNTAX Unsigned32
Hello? Do they really intend to mix Integer32, Unsigned32,
and OCTET STRING in a single object? If so, have they
specified the encodings? (If they really want to do this,
the SMI has a way ... it's called Opaque ... not politically
correct, but IMHO the right thing to do if you are going to
be mising types like this.)
> TeHopAddressIPv4 ::= TEXTUAL-CONVENTION
> DISPLAY-HINT "1d.1d.1d.1d"
> STATUS current
> DESCRIPTION
> "Represents an IPv4 network address."
> SYNTAX OCTET STRING (SIZE (4))
>
> TeHopAddressIPv6 ::= TEXTUAL-CONVENTION
> DISPLAY-HINT "2x:2x:2x:2x:2x:2x:2x:2x%4d"
> STATUS current
> DESCRIPTION
> "Represents an IPv6 network address."
> SYNTAX OCTET STRING (SIZE (16))
These look a lot like clones of InetAddressIPv4 and
InetAddressIPv6 from from RFC 3291.
> TeHopAddressLspID ::= TEXTUAL-CONVENTION
> DISPLAY-HINT "1d.1d.1d.1d:2d"
> STATUS current
> DESCRIPTION
> "Represents a unique ID for a CR-LDP LSP. This ID is
> assigned by the head end LSR, and consists of an IPv4
> address belonging to the head end followed by a two
> octet unsigned integer that is unique for each LSP that
> starts at this head end."
> SYNTAX OCTET STRING (SIZE (6))
>
> TeHopAddressType ::= TEXTUAL-CONVENTION
> STATUS current
> DESCRIPTION
> "A value that represents an address type for a Tunnel
> hop, taken from the following list:
>
> unknown(1) An unknown address type.
> ipv4(2) An IPv4 network address.
> ipv6(3) An IPv6 network address.
> asnumber2(4) A two octet Autonomous System number.
> asnumber4(5) A four octet Autonomous System number.
> unnum(6) An unnumbered interface index.
> lspid(7) An LSP ID, for CR-LDP Tunnels [RFC3212].
>
> Each definition of a concrete TeHopAddress value must
> be accompanied by a definition of a textual convention
> for use with that TeHopAddressType."
> SYNTAX INTEGER {
> unknown(1),
> ipv4(2),
> ipv6(3),
> asnumber2(4),
> asnumber4(5),
> unnum(6),
> lspid(7)
> }
>
> TeHopAddressUnnum ::= TEXTUAL-CONVENTION
> DISPLAY-HINT "d"
> STATUS current
> DESCRIPTION
> "Represents a 32bit unnumbered interface index."
> SYNTAX Unsigned32
Presented out of order, and another Unsigned32 mixed with OCTET STRING.
> Thanks,
> Bert
How can you stand it?
//cmh