EDI Codes and Routing

Introduction

An EDI Code is analogous to an address found on the envelope of a letter and forms the basis for all EDI routing. In this section we look at the use of EDI codes. There are two basic types of EDI code: physical codes and logical codes. A physical EDI code pertains to a physical computer or EDI system. A logical EDI Code could be:

• a computer application such as an invoicing system
• a trading sub division of a company, for instance a manufacturing plant
• a company (in legal terms)
• a corporation (a collection of companies)

EDI Routing

EDI routing, the path which guides the EDI data to its final destination, may be a long and winding trail. Organisations themselves are often complex and a large multinational organisation may hide a world-wide network with many computers and systems. Files entering through a corporate gateway (a single access point from the corporation to the outside world) must be onward routed by that organisation to their final destination.

There are three levels at which EDI data is routed through such a system.

• Physical routing (in OFTP terms called SSID routing). This is routing during a communications connection by physically contacting another system using a network such as the X.25 packet switching system. The EDI code at this level is referred to as the SSID code.
• File routing (in OFTP terms called SFID routing). The largest collection of data crossing a physical connection will always be a file. A file may contain many EDI interchanges but will still always be considered as a single file. The EDI code at this level is referred to as the SFID code.
• Interchange routing (sometimes called UNB, STX or ISA routing). Examination of the contents of an EDI file will enable the forwarder to split a single file into one or more EDI interchanges, all of which will contain interchange control segments with origin and destination EDI codes. The EDI code at this level is referred to as the EDI code.

The first of these three routing methods handles the physical routing and the latter two methods handle logical routing.

Routing Nodes

The entities which exist at the three levels of routing are sometimes called nodes. The three basic node types are:

• Network Node (sometimes called the Physical Node or SSID Node)
• File Node (sometimes called the SFID Node)
• Message Node (sometimes called the UNB Node)

The last two node types, File and Message Nodes, are also called Logical Nodes. The Message Node is not strictly part of OFTP addressing.

Each Network node may have one or more File nodes and each File node may have one or more Message nodes.

Illustration of EDI Routing

A simple analogy for EDI routing nodes (origin or destination points of contact) is to consider the structure of a medium to large company. Incoming mail for that company will be delivered to the mailroom. The mailroom is therefore the equivalent of the network node as it is the point of contact with the outside world.

The mail will then be sorted and delivered to each department in bundles. The department is the equivalent of the file node.

Finally each letter will be delivered to an employee, or section, within that department. This is the equivalent of the message node and the information has reached its final destination.

In smaller companies some of these sections will be omitted so that the mailroom, the department and the final destination are all the same person. So it may be with EDI. In this case all three levels will share the same EDI code.

The Hierarchy of Node Types

To further understand the reason for these different types of node and hence the attributes of each node, it is necessary to look at the communications aspect of an EDI file.

The Session Level

Let us assume that an EDI file has been scheduled to be sent to the Accounts Department’s Sales Ledger section of company A. A communications session is started with company A. The first contact and first level of validation that this data is being sent to the correct place is when the trading partners identify each other and exchange passwords. This level of communication is carried out at Network Node level. In the OFTP protocol, the SSID (Start Session Identity) is used to exchange this information and agree certain standards between the partners.

The File Level

Once the session has been started and passwords have been exchanged, the file may be sent. High level information is exchanged and again there has to be agreement between the two partners that, for example, the receiving partner has an Accounts Department (the destination of the file) and enough room on his system for the file to be sent. In the OFTP protocol, the SFID (Start File Identity) is used to exchange this information.

The Message Level

Finally the EDI data itself may now be transferred between the partners. The EDI data contains its own internal addressing. In the UNB segment (for EDIFACT), the STX segment (for UNGTDI) or ISA segment (for ANSI X12), there must be routing information defining both the sender’s and receiver’s EDI Codes. In this case the EDI code of the Sales Ledger section will be defined in the message as the receiver of this data. In practice, many users, particularly smaller companies, do not wish to use all three levels of routing. These users will have only one EDI code defining all three levels. ODEX can cater for these users and trading partners as well by allowing all three levels to be defined in a single user-profiling operation, creating a combined Network Node, File Node and Message Node with the same EDI Code.