Re: Addendum to WFD byte limit increase

Erik Katsavounidis (Erik.Katsavounidis@lngs.infn.it)
Thu, 10 Apr 1997 20:04:16 +0200 (CET-DST)

>>
>> Hi--
>>
>> The 60K WFD readout that was introduced in MACRO 24 hours ago seems to
>> be affecting acquisition spy jobs by reducing their efficiency. This is
>> due to the fact that more ID=2 are now created. No observable overall
>> data volume increase was seen over the last 24 hours; this suggests that
>> the 60K is really marginal in creating an ID=2 buffer for every ID=1. It
>> was actually Jan Reynoldshon that first noticed a drop in the online
>> trigger rates and hinted that it might be related to the byte limit increase.
>> Off-line, there's nothing observable...
>>
>
>But we have seen unexplainable drops in trigger rates before. Are we
>sure there is a correlation with the BYTE limit increase?

The acquisition trigger rates as *reported* (only, no actual trigger was
actually lost from logging to the data file) by the online system were
systematically down to 30% their nominal values. This applies for all the
trigger all the time. The decrease of the byte limit to 50K brought these
values back to their nominal.

>> For the moment we decided to reduce the WFD readout byte limit to 50K for now
>> (RUN13796) which is still 10K on top of the 40K limit we had since last summer.
>> The main concern is the GC on-line analysis. Kate and Alec, do you have means of
>> checking the buffer collection efficiency of the ERP on-line analysis? Did you
>> see any drop over the last 24 hours? Please let me know.
>>
>
>Actually, since I am beeper sitting for Kate I may have a preliminary answer
>here. During the run where the byte limit was increased (RUN013791)
>there was a GC alarm. (6 hits over energy threshold with a
>coincidence threshold of 5 hits in a 6sec period). So I would be
>surprised if the number of online gc buffers went down, though I am
>not an expert so I can't make a stronger statement.
>
>> We expect to fine tune the WFD byte limit value when the final WFD fix will
>> be implemented.
>>
>
>Based on the measurments Ioannis and I did 2 weeks ago, we didn't need
>much more than 10usec extra.
>
>-Chris
>
>
--Erik