Use the traceon command to enable the trace facility. The trace facility enables you to capture information to assist in the diagnosis of Connect:Direct® problems.
The format for the traceon command follows:
|[filesize=nnK | nnM | 0]|
|[level=basic | intermediate | full]|
|[type=cmgr | pmgr | smgr | comm | (list)]|
|[wrap=yes | no]|
|[dest=destination | (list)]|
|[pnode | snode]|
|[pname=process name | (list)]|
|[pnum=process number | (list)]|
Specify one or more of the following parameters to define the traces:
|file=filename||The name of the output file if you want to write the trace information to a file. The size of the name can range from 1 to 256 characters. The default is CDTRACE.CDT in the Connect:Direct directory.|
|filesize=nnnK | nnnM | 0||The maximum file size as a number of kilobytes (K) or megabytes (M). A value of 0 indicates that the file can grow indefinitely.|
|level=basic | intermediate | full||
The level of trace detail.
basic—Produces a trace of function entry and exit.
intermediate—Produces a trace of function entry and exit, plus arguments to functions.
full—Produces a trace with function entry/exit, function argument display, plus display of selected internal control blocks.
|type=cmgr | pmgr | smgr | comm | (list)||
The type of event or a list of events to be traced.
cmgr—Traces events relating to the interaction of the server with connected workstations and the server console.
pmgr—Traces events relating to the manipulation of Connect:Direct Processes. This trace type provides information about the submission, update, deletion, selection for processing, and termination of Connect:Direct Processes.
smgr—Traces events relating to the execution of Connect:Direct Processes and the server's interaction with other Connect:Direct nodes. Traces can be started for specific Process numbers or names, for specific destinations, or for all PNODE or SNODE Session Managers.
comm—Traces only interactions with external communications facilities invoked from Session Manager threads and used to communicate with other Connect:Direct nodes. You can start traces for specific Process numbers or names, for specific destinations, or for all PNODE or SNODE Session Managers.
|wrap=yes | no||
Specifies whether you want your trace output to be a wraparound file.
no—Requests a linear file that grows until either the user–specified space restriction is reached or you run out of disk space. Linear traces are useful when you can qualify the trace criteria sufficiently or when it is important to not miss any trace events.
yes—Requests a wraparound file that, when the user–specified space restriction is reached, wraps back to the beginning and overwrites the oldest records. Wraparound traces are useful when the problem you are trying to trace occurs intermittently or is very difficult to reproduce.