Teledyne LeCroy SAS Suite User Manual - Manual (Page 96)

Version 6.25 SASSuite User Manual
86
LeCroy Corporation
Recording Size
With Spooled Recordings, you are given the option of setting the recording length based
on time or on the recording size.
Record for - Enter the duration of the recording in days, hours, minutes, and seconds.
Record - Enter the amount of traffic (in MB) that you want the analyzer to record.
Selecting this option will create a fixed length recording that begins as soon as you click
the REC button on the menu bar.
Until Triggered, plus - Enter the amount of traffic (in time) that you want the analyzer to
record following an event trigger. Recording time units are days, hours, minutes, and
seconds.
Until Triggered, plus - Enter the amount of traffic (in MB) that you want the analyzer to
record following an event trigger.
You select event triggers in the Recording Rules page -- see “Recording Rules” on
page 88 for details.
Spooled Recordings - Performance Issues
During a spooled recording if the analyzer buffer fills faster than the trace can be
uploaded to the host PC, the analyzer will briefly suspend recording until some of the
buffer is free. When recording is suspended, gaps will appear in the trace. These gaps
will appear as entries in the trace.
During the recording, you can see if gaps are likely to happen by reading the status bar
at the bottom of the screen.
The Status bar has two fields that you can use to determine if gaps are occurring or are
about to occur.
Gaps - Shows how many gaps have occurred. In the example above, two gaps have
occurred.
Hardware Buffer - shows you how full the analyzer buffer is and if the analyzer is
currently dropping packets. In the example above, Hardware Buffer field shows that the
buffer is full (shown by the thick green graph) and that the analyzer is currently dropping
packets (shown by thin red line atop of the green graph.)
Preventing Gaps
Gaps can be caused by a number of factors - the number of channels being recorded,
the absence of filtering, the performance of the host PC (for example, is it using USB 2.0
to upload traffic?), and the amount of traffic produced by the devices under test.
There are several ways to prevent gaps. You will need to experiment with your setup to
determine what works best for you. You might try filtering out primitives, turning on data
truncation (for example to 10 DWords) and filtering out LISMs. If you want to record
primitives, it is recommended that you enable primitive packing.