Steps in a transformation session and sys.dat file

The table below shows which combinations are possible and which functions are enabled by each combination. The square brackets [ ] indicate an optional function.

Step Code Processing


Read +[Restructure (*)] +[Sort (*)] + [Aggregate Input-Events [by group] (*)]


Read +[Restructure (*)] +[Sort (*)] + [Aggregate Input-Events [by group] (*)] + Transform + [Aggregate Output-Events all groups (*)]


Read + [Restructure (*)] + [Sort (*)] + [Aggregate Input-Events by group (*)] + Transform + [Aggregate Output-Events by group (*)] + [Aggregate Output-Events (*)] + Routing


Transform + [Aggregate Output-Events by group (*)]


Transform + [Aggregate Output-Events by group (*)] + [Aggregate Output-Events (*)] + Routing


[Aggregate Output-Events (*)] + Routing

(*) Not available in MQSeries and JMS modes

Each execution within a session applies the processing specified by the Step keyword in sys.dat.

Successive executions must follow the logical sequence of the three steps.

Execution 1 Execution 2 Execution 3

Step = E

Step = T

Step = S

Step = E

Step = TS


Step = ET

Step = S


Step = ETS


The following diagram shows the processing flow through the steps in the session.

Figure 1: Processing flow

Note   Within a step, you can use keywords that are suffixed with the step letter.
For example, during the E (Input) step, you can use E_xxx keywords.

File Mode

A session requires you to activate at least one of the three possible steps:

  • INPUT (E)
  • OUTPUT (S)

You can activate each of these three steps independently in different executions of a session or combine them in the same execution. The possible combinations of these steps, and how to configure them in sys.dat, are explained in General Options in script.ges.

MQSeries mode and JMS mode

Rule Engine MQSeries or JMS require you to run all three steps (E, T, and S) together.

Related Links