NSAP DSP guidelines: Guidelines for the specification of the structure of the Domain Specific Part (DSP) of the ISO standard NSAP address
How the free part of an ISO network address should be laid out so that addresses can be handed out efficiently.
Braun, H.W., "Guidelines for the specification of the structure of the Domain Specific Part (DSP) of the ISO standard NSAP address", RFC 982, DOI 10.17487/RFC0982, April 1986, <https://www.rfc-editor.org/info/rfc982>.
@misc{rfc982,
series = {Request for Comments},
number = 982,
howpublished = {RFC 982},
publisher = {RFC Editor},
doi = {10.17487/RFC0982},
url = {https://www.rfc-editor.org/info/rfc982},
author = {H.W. Braun},
title = {{Guidelines for the specification of the structure of the Domain Specific Part (DSP) of the ISO standard NSAP address}},
pagetotal = 11,
year = 1986,
month = apr,
} RFC text
Contents
Network Working Group ANSI
Request for Comments: 982 April 1986
Guidelines for the Specification of the Structure of the
Domain Specific Part (DSP) of the ISO Standard NSAP Address
ANSI Working Document X3S3.3/85-258
Status of This Memo
The following is a RFC edition of the ANSI "Guidelines for the
Specification of the Structure of the Domain Specific Part (DSP) of
the ISO Standard NSAP Address". It should be understood that this
ANSI working document is in a draft state and subject to change.
These Guidelines are an important prerequirement to allow IP
addressing within the ISO connectionless protocol (CLNS), which
itself is necessary to allow the coexistance of the IP and the CLNS
in the Arpanet attached gateways.
A RFC to specify a suitable IP addressing method for ISOgrams is
currently worked on in Internet Task Forces and should be expected as
an RFC shortly.
This RFC is for informational purposes only and it's distribution
is unlimited. It does not specify a standard for the Arpa-Internet.
Hans-Werner Braun (HWB@GW.UMICH.EDU)
University of Michigan
ANSI Guidelines
TITLE: GUIDELINES FOR THE SPECIFICATION OF THE STRUCTURE OF THE
DOMAIN SPECIFIC PART (DSP) OF THE ISO STANDARD NSAP ADDRESS
STATUS: X3S3.3 WORKING DRAFT
Introduction
There is a substantial amount of interest in providing guidance to
private address administration authorities on preferred formats and
semantics for the Domain Specific Part (DSP) of an NSAP address. In
particular it is felt that use of a limited number of preferred DSP
formats would allow organizations to easily use intermediate and end
systems from a variety of suppliers. This technical report specifies
the way in which the DSP may be constructed so as to facilitate
efficient address assignment.
1. Scope and Field of Application
This Technical Report specifies a recommended syntax and
interpretation for the Domain Specific Part of an NSAP address. NSAP
Address Administration Authorities are urged to use this format when
the equipment comprising an addressing domain is of a heterogeneous
nature.
2. References
ISO 6523 Data Interchange - Structure for the
identification of Organizations
ISO 7498 Information Processing Systems - Open Systems
Interconnection - Basic Reference Model
ISO 8348 Information Processing Systems - Data
Communications - Network Service Definition
ISO DIS 8348/DAD2 Information Processing Systems - Data
Communications - Addendum to the Network
Service Definition Covering Network Layer
Addressing
ISO DIS 8802 Information Processing Systems - Data
Communications - Local Area Networks
ECMA TR/20 Layer 4 to 1 Addressing
ANSI Guidelines
3. Definitions
(This section to be suppled)
4. Abbreviations
(This section to be supplied)
5. Addressing Principles
This section provides a summary of the major points contained in DIS
8348/DAD2, as they are relevant to this technical report.
Network addresses are defined to be hierarchical. An authority may
either assign a complete address, or else may identify a subdomain of
its own addressing domain within which addresses may be further
assigned by an identified authority for the subdomain. This is done
in such a way that all addresses are unique.
When an authority identifies a subdomain, this creates in effect a
prefix which applies to all addresses assigned within the subdomain.
Certain methods of assigning authority are recognized within the body
of DIS 8348/DAD 2. These take account of existing addressing
standards such as X.121. The abstract structure of an address is
shown in Figure 1.
ANSI Guidelines
+---------------------------+---------------------------+
I I I
I I D P I D S P I
I I I
+---------------------------+---------------------------+
: :
: :
V V
+-----+---------------------+
I I I
I AFI I I D I I
I I I
+-----+---------------------+
Figure 1: NSAP Address Structure
The components of this structure are as follows.
IDP: Initial Domain Part
This is the part of the address whose entire content is
standardized. It consists of two parts, the AFI and the IDI.
AFI: Authority and Format Identifier
This specifies the format of the IDI, the authority responsible
for allocating IDI values, and the syntax of the DSP.
IDI: Initial Domain Identifier
This identifies the subdomain from which DSP values are allocated,
and the authority responsible for allocating the values.
DSP: Domain Specific Part
The contents and semantics of the DSP are not specified in DIS
8348/DAD2. Its abstract syntax is however defined by the AFI. In
most cases, this means that the DSP abstract syntax can be either
binary or decimal.
IDI formats specified in DIS 8348/DAD2 identify a number of
authorities for IDI allocation. These are:
1. X.121 (Public data network numbering)
2. F.69 (Telex numbering)
ANSI Guidelines
3. E.163 (Telephone numbering)
4. E.164 (ISDN numbering)
5. ISO DCC (Geographical address assignment under ISO authority.
A subset of the DCC values used in X.121 have been used to
designate the national addressing authorities)
6. ISO 6523-ICD (Non-geographic address assignment under ISO
authority, using the International Code Designators according
to ISO 6523)
7. Local (the IDI is null and the entire address is contained in
the DSP).
6. Analysis
(This section to be supplied)
7. Recommendations
7.1 X.121 DSP format guidelines
Description of Binary syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. If DSP is present, a DSP of 2 or less octets is an NSAP
selector. A DSP between 3 and 9 octets is interpreted in the
manner described below.
The length of the subnetwork identification is fixed at two
octets (16 bits). The assignment of subnetwork
identifications is under the control of the administering
authority for the private domain, except that the value with
both octets hex "FF" is reserved to indicate that the
subnetwork is not explicitly identified.
The length of the subnetwork address may be up to six octets,
and is determined by the DSP length minus three octets. For
subnetworks using DIS 8802 MAC addressing, this field consists
of the octets of the MAC address in the same order as they are
defined in DIS 8802.
The NSAP selector occupies a single octet.
ANSI Guidelines
Description of Decimal syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. If DSP is present, a DSP of 8 or less decimal digits is
an NSAP selector. A DSP between 8 and 23 decimal digits is
interpreted in the manner described below.
The length of the subnetwork identification is fixed at five
decimal digits. The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that a value of "99999" is reserved to
indicate that the subnetwork is not explicitly identified.
The length of the subnetwork address may be up to fifteen
decimal digits, and is determined by the DSP length minus eight
decimal digits.
The NSAP selector occupies three decimal digits.
7.2 ISO DCC DSP format guidelines
Description of Binary syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. The length of the DSP is between 6 and 12 octets.
The first three octets identify an organization which has been
designated as an addressing authority. This authority is
responsible for assigning values to the remainder of the DSP.
The length of the subnetwork identification is fixed at two
octets (16 bits). The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that the value with both octets hex "FF"
is reserved to indicate that the subnetwork is not explicitly
identified.
The length of the subnetwork address may be up to six octets,
and is determined by the DSP length minus three octets. For
subnetworks using DIS 8802 MAC addressing, this field consists
of the octets of the MAC address in the same order as they are
defined in DIS 8802.
The NSAP selector occupies a single octet.
ANSI Guidelines
Description of Decimal syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. The length of the DSP is between 14 and 29 decimal
digits.
The first six digits identify an organization which as been
designated as an addressing authority. This authority is
responsible for assigning values to the remainder of the DSP.
The length of the subnetwork identification is fixed at five
decimal digits. The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that a value of "99999" is reserved to
indicate that the subnetwork is not explicitly identified.
The length of the subnetwork address may be up to fifteen
decimal digits, and is determined by the DSP length minus eight
decimal digits.
The NSAP selector occupies three decimal digits.
7.3 F.69 DSP format guidelines
Description of Binary syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. If DSP is present, a DSP of 2 or less octets is an NSAP
selector. A DSP between 3 and 9 octets is interpreted in the
manner described below. Interpretation of a DSP greater than 9
octets in length is not specified in this report.
The length of the subnetwork identification is fixed at two
octets (16 bits). The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that the value with both octets hex "FF"
is reserved to indicate that the subnetwork is not explicitly
identified.
The length of the subnetwork address may be up to six octets,
and is determined by the DSP length minus three octets. For
subnetworks using DIS 8802 MAC addressing, this field consists
of the octets of the MAC address in the same order as they are
defined in DIS 8802.
ANSI Guidelines
The NSAP selector occupies a single octet.
Description of Decimal syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. If DSP is present, a DSP of 8 or less decimal digits is
an NSAP selector. A DSP between 8 and 23 decimal digits is
interpreted in the manner described below. Interpretation of a
DSP greater than 23 decimal digits in length is not specified
in this report.
The length of the subnetwork identification is fixed at five
decimal digits. The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that a value of "99999" is reserved to
indicate that the subnetwork is not explicitly identified.
The length of the subnetwork address may be up to fifteen
decimal digits, and is determined by the DSP length minute
eight decimal digits.
The NSAP selector occupies three decimal digits.
7.4 E.163 DSP format guidelines
Description of Binary syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. If DSP is present, a DSP of 2 or less octets is an NSAP
selector. A DSP between 3 and 9 octets is interpreted in the
manner described below. Interpretation of a DSP greater than 9
octets in length is not specified in this report.
The length of the subnetwork identification is fixed at two
octets (16 bits). The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that the value with both octets hex "FF"
is reserved to indicate that the subnetwork is not explicitly
identified.
The length of the subnetwork address may be up to six octets,
and is determined by the DSP length minus three octets. For
subnetworks using DIS 8802 MAC addressing, this field consists
of the octets of the MAC address in the same order as they are
defined in DIS 8802.
ANSI Guidelines
The NSAP selector occupies a single octet.
Description of Decimal syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. If DSP is present, a DSP of 8 or less decimal digits is
an NSAP selector. A DSP between 8 and 23 decimal digits is
interpreted in the manner described below. Interpretation of a
DSP greater than 23 decimal digits in length is not specified
in this report.
The length of the subnetwork identification is fixed at five
decimal digits. The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that a value of "99999" is reserved to
indicate that the subnetwork is not explicitly identified.
The length of the subnetwork address may be up to fifteen
decimal digits, and is determined by the DSP length minus eight
decimal digits.
The NSAP selector occupies three decimal digits.
7.5 E.164 DSP format guidelines
Description of Binary syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. If DSP is present, a DSP of 2 or less octets is an NSAP
selector. A DSP between 3 and 9 octets is interpreted in the
manner described below. Interpretation of a DSP greater than 9
octets in length is not specified in this report.
The length of the subnetwork identification is fixed at two
octets (16 bits). The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that the value with both octets hex "FF"
is reserved to indicate that the subnetwork is not explicitly
identified.
The length of the subnetwork address may be up to six octets,
and is determined by the DSP length minus three octets. For
subnetworks using DIS 8802 MAC addressing, this field consists
of the octets of the MAC address in the same order as they are
defined in DIS 8802.
ANSI Guidelines
The NSAP selector occupies a single octet.
Description of Decimal syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. If DSP is present, a DSP of 8 or less decimal digits is
an NSAP selector. A DSP between 8 and 23 decimal digits is
interpreted in the manner described below. Interpretation of a
DSP greater than 23 decimal digits in length is not specified
in this report.
The length of the subnetwork identification is fixed at five
decimal digits. The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that a value of "99999" is reserved to
indicate that the subnetwork is not explicitly identified.
The length of the subnetwork address may be up to fifteen
decimal digits, and is determined by the DSP length minus eight
decimal digits.
The NSAP selector occupies three decimal digits.
7.6 ISO 6523-ICD DSP format guidelines
Description of Binary syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. The length of the DSP is between 5 and 11 octets.
The first two octets identify an organization which has been
designated as an addressing authority. This authority is
responsible for assigning values to the remainder of the DSP.
The length of the subnetwork identification is fixed at two
octets (16 bits). The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that the value with both octets hex "FF"
is reserved to indicate that the subnetwork is not explicitly
identified.
The length of the subnetwork address may be up to six octets,
and is determined by the DSP length minus three octets. For
ANSI Guidelines
subnetworks using DIS 8802 MAC addressing, this field consists
of the octets of the MAC address in the same order as they are
defined in DIS 8802.
The NSAP selector occupies a single octet.
Description of Decimal syntax
The length of the DSP can be determined since the total address
length is known, and the IDP has fixed length for any given AFI
value. The length of the DSP is between 12 and 27 decimal
digits.
The first four digits identify an organization which has been
designated as an addressing authority. This authority is
responsible for assigning values to the remainder of the DSP.
The length of the subnetwork identification is fixed at five
decimal digits. The assignment of subnetwork identifications
is under the control of the administering authority for the
private domain, except that a value of "99999" is reserved to
indicate that the subnetwork is not explicitly identified.
The length of the subnetwork address may be up to fifteen
decimal digits, and is determined by the DSP length minus eight
decimal digits.
The NSAP selector occupies three decimal digits.
7.7 Local DSP format guidelines
No recommendations made.
ISO Plain Language
On this page
§1Scope of the Guidelines
This RFC is an edition of the ANSI working document X3S3.3/85-258, and the working document was in a draft state and subject to change. It specifies a recommended syntax and interpretation for the Domain Specific Part (DSP) of an NSAP address, and it urges NSAP address administration authorities to use this format when the equipment making up an addressing domain is of a heterogeneous nature.
The guidelines exist because IP addressing inside the ISO connectionless protocol (CLNS) needed them, and that in turn was needed to let IP and CLNS coexist in the Arpanet attached gateways. A separate RFC for an IP addressing method for ISOgrams was being worked on in Internet task forces at the time.
The RFC is for informational purposes only and does not specify a standard for the Arpa-Internet.
The memo publishes a draft ANSI document about how the free part of an ISO address should be laid out. It says the layout matters when a network mixes equipment from different suppliers, and it asks address authorities to follow it. It sets no Internet standard.
§5The Hierarchical NSAP Address
Network addresses are defined to be hierarchical. An authority may either assign a complete address, or identify a subdomain of its own addressing domain within which addresses may be further assigned by an identified authority for the subdomain. This is done in such a way that all addresses are unique. When an authority identifies a subdomain, it creates in effect a prefix that applies to all addresses assigned within that subdomain.
An NSAP address has two top-level parts. The Initial Domain Part (IDP) is the part whose entire content is standardized, and it consists of the Authority and Format Identifier (AFI) and the Initial Domain Identifier (IDI). The AFI specifies the format of the IDI, the authority responsible for allocating IDI values, and the syntax of the DSP. The IDI identifies the subdomain from which DSP values are allocated, and the authority responsible for allocating those values. The Domain Specific Part (DSP) holds the part of the address that the authority assigns; its contents and semantics are not specified in DIS 8348/DAD2, but its abstract syntax is defined by the AFI, and in most cases it can be either binary or decimal.
The IDI formats recognized in DIS 8348/DAD2 are X.121 for public data network numbering, F.69 for telex numbering, E.163 for telephone numbering, E.164 for ISDN numbering, ISO DCC for geographical address assignment under ISO authority, ISO 6523-ICD for non-geographic address assignment under ISO authority, and Local, where the IDI is null and the whole address is contained in the DSP.
An NSAP address has a standardized front part and a free back part. The front part names the authority that hands out addresses and says how the back part must be written. The back part is the Domain Specific Part, and this RFC is about how to fill it in.
- NSAP address Figure 1 of RFC 982, with the DSP layout of section 7.1
- Initial Domain Part (IDP) Entirely standardized
- Authority and Format Identifier (AFI)
- Initial Domain Identifier (IDI)
- Domain Specific Part (DSP) Assigned by the authority named in the IDI
- Subnetwork identification 2 octets, or 5 decimal digits
- Subnetwork address Up to 6 octets, or 15 decimal digits
- NSAP selector 1 octet, or 3 decimal digits
- Initial Domain Part (IDP) Entirely standardized
§7.1The X.121 DSP Layout
The X.121 guidelines are the base case, and the other formats follow the same shape. The length of the DSP can be determined because the total address length is known and the IDP has a fixed length for any given AFI value. In binary syntax, a DSP of 2 or less octets is an NSAP selector, and a DSP between 3 and 9 octets is interpreted as described below.
In the binary layout the subnetwork identification is fixed at two octets, 16 bits. Its assignment is under the control of the administering authority for the private domain, except that the value with both octets hex FF is reserved to indicate that the subnetwork is not explicitly identified. The subnetwork address may be up to six octets and is determined by the DSP length minus three octets. For subnetworks using DIS 8802 MAC addressing, this field consists of the octets of the MAC address in the same order as they are defined in DIS 8802. The NSAP selector occupies a single octet.
The decimal syntax follows the same pattern. A DSP of 8 or less decimal digits is an NSAP selector, and a DSP between 8 and 23 decimal digits is interpreted in the same way. The subnetwork identification is fixed at five decimal digits, with the value 99999 reserved to indicate that the subnetwork is not explicitly identified. The subnetwork address may be up to fifteen decimal digits and is determined by the DSP length minus eight decimal digits. The NSAP selector occupies three decimal digits.
The X.121 layout is the template for the rest. The free part of the address holds, in order, a short subnetwork number, a subnetwork address such as a hardware address, and a final selector that picks the service at that address. The number of digits or bytes used by the subnetwork number and the selector is fixed, so the address length tells you how much room the subnetwork address has.
§7.2Layouts With an Addressing Authority Prefix
Two of the IDI formats add a prefix that names the organization acting as an addressing authority, and both fix a total length for the DSP. For ISO DCC in binary syntax, the DSP is between 6 and 12 octets, and the first three octets identify an organization which has been designated as an addressing authority, responsible for assigning the values in the remainder of the DSP. In decimal syntax the DSP is between 14 and 29 decimal digits, and the first six digits identify that organization.
ISO 6523-ICD uses the same pattern with different sizes. In binary syntax the DSP is between 5 and 11 octets, and the first two octets identify the organization designated as an addressing authority. In decimal syntax the DSP is between 12 and 27 decimal digits, and the first four digits identify the organization.
After that prefix, both formats use the same subnetwork identification, subnetwork address and NSAP selector as the X.121 case: two octets for the subnetwork identification in binary, with hex FF reserved, and five decimal digits, with 99999 reserved, in decimal; a subnetwork address of up to six octets or fifteen decimal digits; and an NSAP selector of one octet or three decimal digits.
The ISO DCC and ISO 6523-ICD layouts start with a few digits or bytes that name the organization allowed to hand out the rest of the address. After that prefix they look like the X.121 layout, and the total length of the free part is fixed by the standard.
§7The Other IDI Formats and the Local Case
F.69, E.163 and E.164 use the same layout and the same limits as X.121. In binary syntax, a DSP of 2 or less octets is an NSAP selector, a DSP between 3 and 9 octets is interpreted as described, and the interpretation of a DSP greater than 9 octets is not specified in the report. In decimal syntax, a DSP of 8 or less digits is an NSAP selector, a DSP between 8 and 23 digits is interpreted as described, and interpretation of a DSP greater than 23 digits is not specified. The subnetwork identification, the subnetwork address and the NSAP selector follow the X.121 rules in both syntaxes.
For the Local IDI format, where the IDI is null and the entire address is contained in the DSP, no recommendations are made.
Across every format, two values are reserved: hex FF for a subnetwork that is not explicitly identified, and the decimal 99999 for the same purpose. The guidelines therefore give a small number of preferred DSP formats, which the RFC says allows organizations to use intermediate and end systems from a variety of suppliers and lets addresses be assigned efficiently.
Three more formats, for telex, telephone and ISDN numbers, copy the X.121 layout and its limits, and say nothing about longer free parts. The local format, where the whole address is free, gets no recommendation at all. Only one value is reserved in each syntax, for a subnetwork that has no number of its own.
| IDI format | Binary DSP | Decimal DSP |
|---|---|---|
| X.121 | 2 octets or less is a selector; 3 to 9 octets are laid out | 8 digits or less is a selector; 8 to 23 digits are laid out |
| ISO DCC | 6 to 12 octets | 14 to 29 digits |
| F.69 | 2 octets or less is a selector; 3 to 9 octets are laid out | 8 digits or less is a selector; 8 to 23 digits are laid out |
| E.163 | 2 octets or less is a selector; 3 to 9 octets are laid out | 8 digits or less is a selector; 8 to 23 digits are laid out |
| E.164 | 2 octets or less is a selector; 3 to 9 octets are laid out | 8 digits or less is a selector; 8 to 23 digits are laid out |
| ISO 6523-ICD | 5 to 11 octets | 12 to 27 digits |
| Local | No recommendation | No recommendation |