Using the Remote System Worksheet

The Remote System Worksheet should be filled out for every OSI system you intend to connect to. This INCLUDES your local node.

This worksheet represents parts of the OSI stack. The information is broken into six different sections. These six sections (described below) should be filled out by the network administrator for each remote node.

General Information

This contains basic information about the remote system.

Node Name	What the system is called (uxnode1, hpnode2, etc.). This name should match the name given to the system on network maps.
Vendor/Type	The manufacturer and model number (for example, HP9000/J200).
Routing Information	Indicates whether this node acts as an End System (final destination for traffic), an Intermediate System (router), or Both.

Application Addresses

List all the remote applications you will communicate with.

Application Name	What the application is (FTAM responder, Net Management Agent, etc.).
P-, S-, T-Selectors	Upper layer addresses for the application. If the program runs over Transport, only the T-selector should be specified. For Session and X.400, specify only the S and T-selectors. For all other applications, the P, S, and T-selectors should be used. Indicate a NULL selector value by using "NULL."

Subnetwork Addresses

The subnetwork information is typically transparent to the end user, because the Network Layer (CONS or CLNS) determines this information from the NSAP. In some cases, this mapping must be configured on the system which is why this information is requested.

If you have multiple X.25 or LAN cards, use additional worksheets.

NSel	The Network layer selector (typically 0x01). This field may be left blank if it is not explicitly configured on the remote.
CONS/CLNS/X.25 Subaddr	Indicate what (if any) subaddresses are used by these services over X.25. If subaddressing is not used then use "NULL."
X.121	Provide the X.121 address of the remote system.
802.3/FDDI MAC	This is the Media Access Control (MAC) layer address of the remote LAN card. These are always 6 bytes long and should be specified in hexadecimal.

X.25 Subnetwork Information

This box contains other X.25 configuration items that are important to know to achieve successful interoperability.

ISO 8878 Support	This indicates whether the remote supports the use of the X.25 extended addressing facility, as prescribed by ISO 8878 (Use of X.25 to provide CONS). Systems that do not use this facility should have their NSAP match their X.25 address (X.121 + Subaddress).
Accept Reverse Charging	If true, the vendor allows the call charges to be billed "collect." Typically this is set to "N".
X.25 Standard	This indicates which X.25 capabilities are supported by the system. The 1984 standard provides more facilities than the 1980 standard. At least one must be circled. (Circle all that apply or will be allowed by the X.25 service).

LAN Subnetwork Information

This box indicates what subset of the CLNS protocol is supported.

CLNP Subset

U.S. GOSIP requires the use of the Full subset, with NULL CLNP packets being discarded. This is the default for HP systems. The NULL subset is used only when all destination OSI nodes are physically connected via the same LAN. It provides low overhead at the cost of internetworking. It also requires that the NSAP match the MAC address of the system. Non-segmenting and Full subset (accept NULL) are less frequently used, but are supported by HP.

ES/IS Supported

Indicates whether this node supports the End System/ Intermediate System protocol. With this protocol, subnetwork addressing information is done dynamically. Therefore, other systems on the network do not need to statically configure this addressing information.

Network Addresses (NSAPs)

The last items you should provide are the Network Addresses (sometimes referred to as NSAPs) that are used to access this system.

The full NSAP should be specified here if the remote configures the Network Address in components (IDP, DSP), or does not include the NSEL. These values should all be concatenated and written in the appropriate field.

RFC1006 NSAP and IP Address

OSI uses NSAP network addresses. TCP/IP uses IP addresses. Since the OSI networking layer is replaced by TCP/IP, the special RFC1006 NSAP must be used in any OSI dialog between systems.

The RFC1006 NSAP has:

a specific header with an IP address, for example, RFC1006 NSAP is 540072872203
optionally a TCP port ID appended to it

Therefore, an OSI service running over TCP must specify a destination NSAP using the RFC1006 NSAP structure.

The IP address of 123.55.77.5 is encoded in the RFC1006 NSAP as a string of digits using three digits corresponding to one octet of the IP address.

IP Address of:	Encoded in NSAP as:
123.55.77.5	123055077005

If the default port (102) is not used. the TCP port number is encoded as 5 digits followed by "F".

The complete RFC1006 NSAP with an optional port number of 103 is shown below. If you use the standard port of 102 you omit the port number from the end of the NSAP.

Address Header	IP Address	Port # (optional)
540072872203	123055077005	00103F

If you use the standard NSAP address header, but an invalid IP address, the NSAP address is unreachable.

RFC1006 Addressing Considerations

When addressing a remote, always use the remote's IP address to construct the destination NSAP.
Applications must bind to the RFC1006 NSAP with the IP address as configured in osiadmin whether you are connecting to remote via that address or not. You can also use a wildcard NSAP for the local address.