OSI/TS Configuration and Management Manual

Performing Monitoring and Troubleshooting Guide
OSI/TS Configuration and Management Manual—424831-001
4-8
Some Words of Caution
being used by an particular application), then CHECK SU finds the related NSP
subdevices .
Some Words of Caution
Before using CHECK SU, make sure you have planned for the following:
•
Use the wild card option with great care, especially in a busy OSI network. Wild
cards can have an effect on the performance levels of TSP processes that are
queried. Use of wild cards alone, without a partial name, is not recommended.
•
Use the DETAIL option sparingly because it generates a great quantity of
information. It may be easier to log the output to the spooler or to a disk file and
print out the DETAIL screens.
•
When a connection configured for dynamic subdevices (the DELETETIME attribute
for the SU object is set to 0) terminates, all subdevices created for that connection
are terminated. In this case, CHECK SU wont show anything useful unless the
connection is established. In other words, error information cannot be displayed on
dynamic subdevices if you are having trouble establishing a connection or the
connection has failed.
•
When a connection configured for static subdevices (the DELETETIME attribute
for the SU object is set to -1) terminates, all subdevices opened for that connection
are set to inactive (not deleted). In this case, CHECK SU can display useful
information on failed connection. Since static subdevices are not deleted, error
information for these subdevices is retained by the subsystem and can, therefore, be
retrieved.
•
X25AM returns many different kinds of errors; TLAM/PAM returns very few kinds
of errors. Because TLAM and PAM are very different from X25AM in this regard,
do not assume that TLAM/PAM errors are not registered—they are, indeed,
registered and reported to you. In addition, frames lost on the LAN are indirectly
indicated in the STATUS SERVICE display in the timeout and retransmission count
fields.
•
The ERRORONLY option cannot detect all possible subdevice errors. NSP
subdevice errors usually get reset by the next NSP operation, but are saved in the
Transport Layer and can be displayed by using the CHECK SU TSP command.
TSP subdevice errors, however, get reset by the next I/O operation to the TSP
process and cannot always be saved. Therefore, not all TSP errors can be captured
by CHECK SU.
The ERRORONLY option selects only subdevices that are in an error state and that
are being used by the server specified in the CHECK SU command. All other
subdevices associated with those selected error subdevices are also displayed,
depending upon the selection specifiers included in the command. For
example, if the ERRORONLY option is specified for a TSP process, the CHECK
SU command only looks for TSP subdevices in an error state. Any NSP subdevices
associated with those selected TSP error subdevices are also displayed, whether they
are in an error state or not.