<snip>
 >As said in multiple postings here, it is not acceptable, and not only I'm
happy
 >that LaserRX still works (but I have printing problems - doesn't recognize
 >printers added under Win95 :-(
 >
 >HP seems not to have enough resources (especially CSY lacks of money to
pay the
 >MeasureWare team) to migrate HP3000 Scope into MeasureWare Agent Scope
which
 >would be compatible to PerfView Analyzer on HP-UX. Strange enough that HP
offers
 >this compatible agent for NT, AIX, and a lot of other platforms so that
you have
 >direct access to the Scope files but not for MPE :-(
 >
 >Best regards, Andreas Schmidt, CSC, Germany

<snip>


<venting on>

I feel your pain Andreas.  I'm still running Laser/RX even though I have
access to several NT systems and several copies of PerfView.  For me the
ability of Laser/RX to read the live SCOPE files without an extract is a
HUGE advantage.  My primary use of Laser/RX is to examine short to medium
term problems (1 day to 1 month).  In doing so I find the need to compare
up-to-the-minute statistics of several systems.

  For example say I have a problem that manifests itself every Monday as a
general CPU increase and disc utilization increase.  Normally I would bring
up a bottleneck graph for the previous Monday and the current Monday for
each of the systems I want to check.  Then generate a graph of the current
Monday for each of the systems.  To refresh, I would simply close the
current graphs and reopen them.  The new data would then be displayed.  To
do the same in PerfView would require that I do a manual extract on each
system, download all the extracts to my PC, close the current files, open
the new files and draw the new graphs.  Needless to say, the Laser/RX method
is a lot better than the PerfView method.

  I mentioned this to HP when PerfView first came out and in general I got
the response of "You're miss using Laser/RX. What you really want to use in
such a situation is Glance."  I tried to explain that you can't use Glance
for problems that stretch for hours since it is really designed for problems
that last for seconds or minutes.  It is possible to get some cummulative
statistics, but they are very basic at best.  Apparently no one needs to
investigate performance problems like mine.

  Unfortunately, not even Lund is trying to fill this void.  Perhaps I
really am the only person who has problems like the "cpu/disk increase"
mentioned above.

<venting off>


 ---------------------------------------------------------------
| Michael P. Smith                                              |
| Sr. Systems Programmer                                        |
| The Hertz Corporation    [log in to unmask] (work)          |
| Oklahoma City, OK                                             |
 ---------------------------------------------------------------
The thoughts, views and expressions contained in this message are those
of Michael P. Smith, and do not necessarily reflect those of the Hertz
Corporation.  To reply via email, remove the .xx from email address.