Local database Local testing database seems necessary in current situation A common solution for UCSB/FNAL would be optimal Based in root, ascii file summaries, and html viewable results/plots Similar to BABAR silicon database Results should have interface with official database Upload every week or so Goal is to be able to track cause and amounts of defects, and give suggestions for repair cycle and remove dependence on official database
Root interface First interface with ARCS output root file Generate minimal set of plots which describe data Generate table of test results which allows one to see correlations in failures See BABAR example From table, generate list of bad channels, with most likely cause of failure/repair list and time problem first seen Will need to be able to access sensor database and previous test results
BABAR test table
BABAR summary page
HTML interface Start with plots/lists generated by root analysis software Write PERL (??) script which generates HTML viewable summary of testing results at each stage Write PERL (??) script which generates page per module which shows plots and has links to past results and bad channel list
Ntuple generation/analysis Start with results of root analysis software Generate ntuple of number/location/cause of bad channels Allows for easy generation of plots of number of bad strip generate/ bad strip locations/ etc. Can point of systematic problems, areas for improvement, etc.
Manpower Lisa/Steve both working on forms of root analysis macros New undergraduate (Derek) has PERL scripting experience and some free time with current production rate Hopefully, many routines can be modified from BABAR code I am currently working on bad channel requirements and correlating failed tests with physical problems