COMPASS Online Databases

Part of the DAQ setup is a MySQL RDBMS providing storage for DAQ configuration, run logbook, beamline monitoring, spill statistics and online filter statistics. Running on pccodb01 as primary server only accessible from within the DAQ network (to ensure DAQ stability), the databases are replicated to where they may be accessed more freely.

In the intention to provide a stable interface for online as well as offline use, the different databases and tables are documented here:



This table contains the values of all DAQ scalers, being read out by Cinderella at end-of-burst. The data are intended to be used for online monitoring as well as for offline analysis (e.g. for bad spill list creation).

column nametype explanation
tbname char(8) technical board name of the scaler "plane" which has been read out
eobtime timestamptime when the first subevent of the end-of-burst event was assembled in the LDC
runnb uint run number
spillnb uint spill number
tcscycles uint number of TCS cycles since begin of spill
input1 uint Single scaler values having been read out from the "technical board" given in tbname. The numeric index here denotes the "wire" of the "plane" (not the channel of the CMC). Unused columns are set to zero.
input2 uint
... uint
input30 uint
input31 uint


This table contains the values of the "Munich" scalers, being read out indepently from the DAQ at regular intervals.


Here the counts of accepted events are stored for every spill and every trigger to allow online monitoring of Cinderella.

column name type explanation
eobtime timestamptime when the first subevent of the end-of-burst event was assembled in the LDC
run uint run number
spill uint spill number
pccoeb uint number of the event builder machine
triggernumberuint trigger number (started counting at 1), the value 255 indicates all triggers
accepted uint number of accepted event
statistics uint total number of events

As every Cinderella instance writes its statistics to separate rows, aggregation of these rows is necessary to retrieve a full summary of online filter performance. The recommended query string to poll data points for online monitoring from this table is:

SELECT triggernumber, MIN(eobtime) as time, SUM(accepted)/SUM(statistics) AS acceptratio FROM FilterInfo WHERE eobtime > date_sub(NOW(), INTERVAL 1 HOUR) GROUP BY run, spill, triggernumber;

Last update: 2006/09/24 by
Valid XHTML 1.1