User's Manual
Top-down analysis explanation
The top-down analysis is a troubleshooting strategy that helps you to more ef
ficiently generate
a performance file database.
Begin the top-down analysis by generating a database that contains only the first and the last
performance data files of the week. Limiting the number of performance data files that are
generated helps you to determine which board, RFP, or PP causes problems during the
week.
You can also limit the number of files imported to the database. Copy the directory structure
that contains all performance files to create the database. Delete the performance files that
are not needed by clicking Start on the Windows taskbar and choosing Search > For Files or
Folders on the taskbar of Windows 2000
™
. Use the time and date stamps on files to find the
performance files that must be deleted. Press <CRTL + A> to select the files, and press the
Delete key.
If you cannot determine what components of the system cause problems, try generating a new
database with one data file each day. If you still cannot solve the problem, add more detail, but
generate a new database each time. Generating databases with more details (that is, more
performance files) substantially increases the duration of the import process. Always consider
if there is enough value added in generating more details to compensate for the extra time this
takes.
Select the following items for generating a more detailed performance database:
• one board only
• upm files (contain PP information)
• epm files (contain board and RFP information)
• a combination of the above selection criteria
If you cannot verify exactly which board, RFP, or PP causes the problem, import only the data
files that contain the relevant information for your problem. For example, if an RFP on a specific
board causes the problem, import the epm files of only that board (epm files contain board and
RFP counters and timers). You can now efficiently generate a new, detailed database that
contains only the relevant information.
Statistical Performance Data
Table 51: Statistical Performance Data
Counter Description
1 Indicators, not used
2 Number of page failures
Performance Collection
DMC DECT Fundamentals August 2012 307