NET/MASTER Network Control Language (NCL) Programmer's Guide

Tracing System-Level Messages
Developing System-Level NCL Procedures
106160 Tandem Computers Incorporated 17–29
Step 2. Specifying the Messages to Trace
The second step in tracing system-level messages is to specify the messages to trace by
using the TRACE trace-spec command. The TRACE trace-spec command makes
changes to a set of trace parameter settings, which includes the trace file name and the
messages to be traced. The parameter settings are stored in a trace parameter list
(TPL), one of which is associated with each Operator Control Services (OCS) window.
Changes made to parameter settings in the TPL are not implemented until you start
the trace. The TRACE ALTER and TRACE RESET commands allow you to alter and
reset trace parameter settings.
The TRACE trace-spec command is typically used from the OCS command input line.
For each type of system-level message flow (EMS, LOG, and MSG), you can trace the
following types of messages:
Messages arriving at the system-level NCL procedure: for example, all messages
arriving at EMSPROC from the EMSREAD verb.
Messages deleted by the system-level NCL procedure: for example, all messages
deleted by LOGPROC by the LOGDEL verb.
Messages sent from the system-level NCL procedure: for example, all messages
sent from a MSGPROC by the MSGCONT verb.
All messages arriving at, deleted by, and sent from the system-level NCL
procedure.
To specify that all messages deleted from a MSGPROC NCL procedure are to be
traced, enter the following command from the OCS command input line:
TRACE MSG=DEL RECSIZE=0
Since only messages with text AAAAA or ZZZZZ are deleted by the preceding
MSGPROC NCL procedure, the trace file should contain only these messages.
The RECSIZE operand of the TRACE trace-spec command specifies the amount of text
that is formatted and traced in a message. Specifying 0 (zero) means that the whole
message is traced; that is, no truncation occurs, and the whole message is traced. It is
recommended that you specify 0 (zero) when tracing system-level messages, which are
mapped by the map $MSG. See Section 11, “Standard and User-Defined Maps,” for
the structure of $MSG.
Step 3. Starting the Trace
After you have specified the system-level messages to trace in the TPL, the third step is
to start the trace by using the TRACE START command. This command starts a trace
collector process, which opens a trace file to collect trace records and controls the
writing of trace records to the trace file.
The results of the command include the process name of the trace collector process, a
unique trace ID from 1 through 65535, and the user ID of the user starting the trace.
(You can use the trace ID in subsequent TRACE ALTER or TRACE STOP commands
to identify the trace.)