AutoTMF Software User's Guide (Update 11)
System Management
HP NonStop AutoTMF Software User’s Guide—429952-013
A-20
Executing Prepared Applications
•
A user issues a TRACE command that causes a selected process or processes to 
be traced or enter debug. Some other user starting a process may not know that 
the process is being traced or debugged.
•
A user specifies DEFINEs, such as =_ESCORT_STATEMENT_TRACE when 
executing the process; in this case, the user has control over tracing and 
debugging.
The TRACE command can be secured to prevent exposure, as follows:
•
The global setting for SECURETRACE should be ON; use the ALTER GLOBAL 
command to set the SECURETRACE setting. The SysDB should be suitably 
secured to prevent unauthorized users from altering this setting.
When SECURETRACE is ON, the user who issues the TRACE command must be 
the process access ID of the processes that are to perform tracing. Other 
processes will run normally and will not trace or enter debug.
In development and testing systems, or systems where security is not a major 
issue, you may wish to set SECURETRACE to OFF to facilitate problem analysis.
•
When a user wishes to trace DATA (that is, produce a trace in which the images of 
data records accessed by the program are displayed), NonStop AutoTMF software 
requires the trace file to be secured so that only the process access id of the 
process may read the file. This ensures that no unauthorized person can read the 
database records in the trace.
Executing Prepared Applications
Swap space
When an application process first opens a file, the NonStop AutoTMF software runtime 
allocates an extended memory segment, which requires a disk swap file. Space 
allocation or volume security errors may cause the allocation to fail, which will 
terminate the application process.
The default location for a swap file is the volume on which the object file resides. If the 
space on that volume is insufficient, the user can allocate up to four alternate swap 
volumes using the global parameters SWAPVOL[0],SWAPVOL1, SWAPVOL2, 
SWAPVOL3. See ALTER GLOBAL on page 6-34 for details.
EMS Logging
In addition to being displayed on the home terminal of the application program, the 
internal errors detected by the runtime are logged to an EMS collector. The default 
collector is $0. Logging to an alternate collector can be configured using the global 
parameter EMSCOLLECTOR or, if required on a process by process basis, by using a 
DEFINE. See ALTER GLOBAL
 on page 6-34 and =_ESCORT_EMS_COLLECTOR on 
page B-6 for details. 










