Conference ilbbak::ibi_focus

99.0. "DTM Benchmarking Experience, problems" by MEMV03::COPPERSMITH (It's an allegory...) Sat Jul 30 1988 10:40

    Recently we used DTM 2.2 to drive a benchmark on an 8700.  This system was 
written in FOCUS, and terminal control was affected by FOCUS, FMS and TPU.  

    We used DTM to record test sessions, then play them back later under
controlled circumstances while measuring the system resource requirements.
This tactic was successful, but we encountered the following problems:

	1)  While recording the sessions, FOCUS CRTFORMS (FOCUS Data entry
	    Screens) would freeze while painting, as though expecting input.  
	    Supplying input would allow the screen paint to continue, until
	    the next freeze.  This extra input could the be backed out
	    and the session continued.  FMS screens caused no problems.

	2)  The reproducibility of the session seemed to be questionable.
	    That is, a session recorded on Thursday would fail on Friday.
	    sessions sometimes seemed to lose their place, or some sessions
	    would stop midway with a message that the session file was in an 
	    illegal format.  (No illicit editing took place).  These may
	    have been caused by time dependent issues.  Sessions recorded
	    on a busy day seemed not to work when the system was not
	    heavily loaded.

    Session files are available for inspection, if anyone would like to
explore these problems.  Does anyone know if such behavior might be banished
in V2.3?  Are there any terminal settings that would affect this behavior?

This note will be posted in DTM, FMS, FOCUS and IBI_FOCUS notes conferences.

Thank you in advance for any comments or help,

Curtis Coppersmith
MEMCL1::COPPERSMITH
DTN 232-2246