OSI/FTAM Configuration and Management Manual

FTAM Troubleshooting
OSI/FTAM Configuration and Management Manual—421944-001
6-2
General Strategy
General Strategy
Troubleshooting an FTAM problem begins when you become aware that a problem
exists. Depending on the kind of problem, you first notice it in one of two main ways:
•
Your application does not function correctly and/or returns errors (in the case of a
remote application using the Tandem responder, your application receives diagnostic
messages from the responder).
•
You receive event messages reporting a problem, generated by APLMGR, initiator,
responder, or other processes.
To ensure that you receive as much information as possible about application problems
if they occur, your applications should follow good error-handling strategy. Applications
using the Tandem initiator and the FTAM API should follow the strategy outlined in the
Tandem OSI/FTAM Programming Guide; applications using the services of the Tandem
responder should follow similar strategies, tailored to the remote system’s interface.
Examine the initial evidence presented in errors and event messages. The problem could
be reported by one of several system components: for example, by APLMGR, by an
initiator or responder process, or by underlying Tandem software (such as an OSI
manager, TAPS, TSP, or NSP process or some component of theTandem NonStop
Kernel operating system). In some cases, error codes or operator-message parameters
indicate further that the problem originated in some other system component, such as an
underlying process.
You can then use tools to pinpoint the cause of the problem. Tandem provides several
tools for this purpose (see Troubleshooting Tools and Their Use
on page 6-3). Each tool
provides different information about potential problems.
If you’ve isolated and corrected a problem that caused an FTAM association to be lost,
your application must reestablish the association. In some cases, you must recover files
first.
Typically, troubleshooting is an iterative process. Once you have found a problem, you
correct it; run the application again; recheck the symptoms; and if an error still exists,
use the tools again. You might need to go back and forth between several tools to find
the cause of the problem. The exact sequence in which you examine evidence and use
the tools depends on the problem. Types of Problems
on page 6-12 describes the
common types of problems and gives hints for troubleshooting each type.