IBM OS/390 User Manual
31.4 JES2 Diagnosis
There are some JES2 mechanisms that can be used for problem determination.
•
$TRACE: JES2 internal tracing can be activated via $S TRACE; $T TRACEDEF;
$P TRACE commands. These are typically requested by IBM service
personnel.
$P TRACE commands. These are typically requested by IBM service
personnel.
•
SYMREC: Symptom records are recorded in the LOGREC file for some JES2
internally discovered problems. Use EREP to format these entries.
internally discovered problems. Use EREP to format these entries.
•
HASP088 message: When JES2 terminated abnormally, a HASP088 message
is generated. This multi-line WTO has registers at ABEND, a traceback of
program linkage leading up to the error, and of course the reason for the
termination.
is generated. This multi-line WTO has registers at ABEND, a traceback of
program linkage leading up to the error, and of course the reason for the
termination.
•
Analysis of maintenance level: The maintenance level of a given module as
well as its memory address can be determined via $D MODULE(name). This
typically will be requested by IBM service personnel.
well as its memory address can be determined via $D MODULE(name). This
typically will be requested by IBM service personnel.
31.5 SLIP
SLIP commands are a powerful tool that can be used to either take an SDUMP or
write trace records to GTF. Some of the SLIP commands are PER (program event
recording) such as instruction fetch and successful branch and may cause CPU
degradation depending upon how they are used. SLIP provides additional
functionality beyond what the VSE SDAIDS does. SLIP commands are generally
entered by a system programmer at an operator console. If the SLIP command
causes trace data to be written, then GTF needs to be started for the trace data
to be recorded.
write trace records to GTF. Some of the SLIP commands are PER (program event
recording) such as instruction fetch and successful branch and may cause CPU
degradation depending upon how they are used. SLIP provides additional
functionality beyond what the VSE SDAIDS does. SLIP commands are generally
entered by a system programmer at an operator console. If the SLIP command
causes trace data to be written, then GTF needs to be started for the trace data
to be recorded.
31.6 Performance Tools
RMF is provided with OS/390, in addition there are several vendor performance
monitors.
monitors.
RMF provides three levels of performance monitoring:
•
MONITOR I - records data to SMF that is used for performance and capacity
analysis. Data is formatted via an RMF batch program or by other products
such as SLR or SAS.
analysis. Data is formatted via an RMF batch program or by other products
such as SLR or SAS.
•
MONITOR II - runs under TSO and provides an interactive view of the OS/390
system (CPU, paging, storage, dispatching).
system (CPU, paging, storage, dispatching).
•
MONITOR III - a powerful interactive tool that monitors the OS/390 system for
bottlenecks in system throughput and provides a point and shoot view of
overall and subsystem performance. Current interval and historical interval
data is provided.
bottlenecks in system throughput and provides a point and shoot view of
overall and subsystem performance. Current interval and historical interval
data is provided.
31.7 LOGREC
The LOGREC data set is functionally equivalent to the VSE recorder file. In
addition to hardware errors OS/390 records software errors to this data set. This
provides valuable information in the initial problem determination phase of
diagnosing a problem. EREP is normally used to extract the required records. In
addition to hardware errors OS/390 records software errors to this data set. This
provides valuable information in the initial problem determination phase of
diagnosing a problem. EREP is normally used to extract the required records. In
Chapter 31. Diagnosing System Problems
475