Dom Test in 2005 Martin Kestel PSU Collaboration

Dom. Test in 2005 Martin Kestel, PSU Collaboration Meeting March 20 2005 Berkeley M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20

Outline • • • System Overview Hardware Requirements Shopping List Readiness Dom. Test Results Summary & Outlook M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 2

DOM a 23 b 7 DOM a 23 bc DOM a 23 bf System Overview Test. DAQ data D F L Data Storage read Pass/ Fail, Dom. Test Char Ped. Pattern Fat Dark. Noise Fat Gain. Vs. HV Fat Time. Resolu Fat ATWDLinea Fat DOMLineari …more to come Gain, Rate, … read Data Base User Inter Plot & Histogram Files PASS read File System Face DOM a 23 bf Detailed Report result 0 =… M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 result 1 = … 3

Hardware Requirements 1/4 Test. DAQ data read D F L Data Storage Pass/ Fail, Dom. Test Char Ped. Pattern Fat Dark. Noise Fat Gain. Vs. HV Fat Time. Resolu Fat ATWDLinea Fat DOMLineari …more to come Gain, Rate, … read Data Base User read File System Ca. 3 TB for all 5000 DOMs to be tested (*) see Appendix PASS Inter Plot & Histogram Files DOM a 23 b 7 DOM a 23 bc DOM a 23 bf Face DOM a 23 bf Detailed Report result 0 =… M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 result 1 = … 4

Hardware Requirements 2/4 Test. DAQ data D F L Data Storage read Pass/ Fail, Dom. Test Char Ped. Pattern Fat Dark. Noise Fat Gain. Vs. HV Fat Time. Resolu Fat ATWDLinea Fat DOMLineari …more to come Recommend fast machine with sufficient RAM; 1 machine per site, PSL: maybe two Gain, Rate, … read Data Base PASS User Inter Plot & Histogram Files DOM a 23 b 7 DOM a 23 bc DOM a 23 bf read File System Face DOM a 23 bf Detailed Report result 0 =… M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 result 1 = … 5

Hardware Requirements 3/4 Test. DAQ data D F L Data Storage read Pass/ Fail, Dom. Test Char Ped. Pattern Fat Dark. Noise Fat Gain. Vs. HV Fat Time. Resolu Fat ATWDLinea Fat DOMLineari …more to come Currently: Desktop PC 2. 4 GHz, 2 GB RAM with 250 GB USB drive Gain, Rate, … read Data Base PASS User Inter Plot & Histogram Files DOM a 23 b 7 DOM a 23 bc DOM a 23 bf read File System Face DOM a 23 bf Detailed Report result 0 =… M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 result 1 = … 6

Hardware Requirements 4/4 Test. DAQ data D F L Data Storage read Pass/ Fail, Dom. Test Char Ped. Pattern Fat Dark. Noise Fat Gain. Vs. HV Fat Time. Resolu Fat ATWDLinea Fat DOMLineari …more to come Prototype available. Runs on any machine with access to database and file system. Gain, Rate, … read Data Base PASS High Level Summary Inter Plot & Histogram Files User DOM a 23 b 7 DOM a 23 bc DOM a 23 bf read File System Face DOM a 23 bf Detailed Report result 0 =… M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 result 1 = … 7

Shopping list • Data Storage: 3 TB over entire IC lifetime • Dom. Test processing: one machine per site (PSL: might need two or dual CPU) (3 GHz CPU, 2 GB RAM min. ) • One database server per site (est. 10 GB/yr) • One file system per site ~100 GB/yr M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 8

DOM a 23 b 7 DOM a 23 bc DOM a 23 bf Readiness 1/5 Test. DAQ data D F L Data Storage read Pass/ Fail, Dom. Test Char Ped. Pattern Fat Dark. Noise Fat Gain. Vs. HV Fat Time. Resolu Fat ATWDLinea Fat DOMLineari …more to come Gain, Rate, … read Data Base User Inter Plot & Histogram Files PASS read File System Will use Ice. Cube metadata server Face DOM a 23 bf Detailed Report result 0 =… M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 result 1 = … 9

DOM a 23 b 7 DOM a 23 bc DOM a 23 bf Readiness 2/5 Test. DAQ data D F L Data Storage read Pass/ Fail, Dom. Test Char Ped. Pattern Fat Dark. Noise Fat Gain. Vs. HV Fat Time. Resolu Fat ATWDLinea Fat DOMLineari …more to come Gain, Rate, … read Data Base User Inter Plot & Histogram Files PASS read File System Framework and processing interface: READY Initial suite of Dom. Tests: currently being fine-tuned More tests to be added; Proc. times in Appendix. M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 Face DOM a 23 bf Detailed Report result 0 =… result 1 = … 10

DOM a 23 b 7 DOM a 23 bc DOM a 23 bf Readiness 3/5 Test. DAQ data D F L Data Storage read Pass/ Fail, Dom. Test Char Ped. Pattern Fat Dark. Noise Fat Gain. Vs. HV Fat Time. Resolu Fat ATWDLinea Fat DOMLineari …more to come Gain, Rate, … read Data Base User Inter Plot & Histogram Files PASS read File System Database: Ready, fine tuning underway. (Recent speed gain by a factor Face DOM a 23 bf Detailed Report result 0 =… of 15 (Dave. G)) File System: an ordinary file system M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 result 1 = … 11

DOM a 23 b 7 DOM a 23 bc DOM a 23 bf Readiness 4/5 Test. DAQ data D F L Data Storage read Pass/ Fail, Dom. Test Char Ped. Pattern Fat Dark. Noise Fat Gain. Vs. HV Fat Time. Resolu Fat ATWDLinea Fat DOMLineari …more to come Currently being fine tuned. Future upgrade: unified, web based interface to STF, Dom. Cal and Dom. Test Databases. Gain, Rate, … read Data Base User Inter Plot & Histogram Files PASS read File System Face DOM a 23 bf Detailed Report result 0 =… M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 result 1 = … 12

DOM a 23 b 7 DOM a 23 bc DOM a 23 bf Readiness 5/5 Test. DAQ data D F L Data Storage read Pass/ Fail, Dom. Test Char Ped. Pattern Fat Dark. Noise Fat Gain. Vs. HV Fat Time. Resolu Fat ATWDLinea Fat DOMLineari …more to come Summary sheet: Design ready (next slide). Detailed Report: webpage with many details READY (*) Gain, Rate, … read Data Base User Summary Sheet Inter Plot & Histogram Files PASS read File System Face DOM a 23 bf Detailed Report result 0 = … M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 result 1 = … 13 (*) See Appendix

Summary Sheet Layout M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 14

Dom. Test Results • Two-Page summary plots are available at the end of each FAT cycle. Can include STF, Dom. Cal results in the same step. • PSL and Dom. Test results agree reasonably well – discrepancies are understood / addressed • Example Detail plots (Fat. Dark. Noise) available in Appendix (and on the web) M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 15

Summary • Dom. Test framework, core set of tests, processing scripts, database IO: READY for 2005 FAT season • Recording test. DAQ data and storing it crucial for Ice. Cube – Always the high likelihood that we will want to reprocess the data • User interface to processing is easy to use • Command line based database extraction READY – Future upgrade proposed: unified, web based interface to STF, Dom. Cal and Dom. Test databases M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 16

Outlook (1/2) Next Steps: • Time Resolution test finalisation • DOM Linearity test finalisation • testlab. Description database launch imminent • domprodtest: add testdaq-io / framework version IDs • webpage: new portal to be filled with content from older domtest portal attempts • MK: Visit to each DFL site, setting up domtest automated processing, databases, etc. (next slide) • Further use follows: pedestal upgrade a la Dave Seckel study secondary pulses at 20 ns, 60 ns calibration approach ATWD ticks/m. V …. . and more M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 17

Outlook (2/2) Travel Schedule Madison, Zeuthen, Uppsala: Beginning and End of FAT dry run and first FAT cycles 2005, ensuring functioning and that local operators are comfortable with Dom. Test: Beg. PSL FAT dry run: Mar 28 thru Mar 31 PSL End Dry Run, Beg. FAT run: Apr 10 thru Apr 18 PSL End FAT run: Apr 25 thru Apr 30 PSL Beg. FAT run: May 10 thru May 13 DESY Beg. FAT run: May 14 thru May 18 Uppsala End FAT run: May 28 thru Jun 02 DESY End FAT run: Jun 03 thru Jun 6 Uppsala Dates are subject to optimization and change M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 18

A 1: Processing Time Estimates for one set of [#Runs]: PSL FAT 3 Pedestal Noise Time. Res Gain #Runs 2 1 7 per DOM: 10" 17" 2'09" 50 DOMs 8'20" 13'59" 1 h 47'54" 100 DOMs 16'40" 27'59" 3 h 35'49" Processed on one CPU of a 3. 4 GHz dual-CPU opteron cluster node with 8 GB RAM M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 19

A 2: Disk Space Estimates Disk space estimates for one entire FAT cycle ============================== Temperatures: +25, -45, -25, +25, -45, +25 C. (neglecting PSL's additional -55 C) Assumption of recording at each temperature: Pedestal: 2 (ATWD 0, ATWD 1) Noise: 2 (ATWD 0, ATWD 1 separately) Gain: 14 runs (ATWD 0, ATWD 1 separately) Time. Res: 4 (ATWD 0, ATWD 1 separately, two laser frequencies) Linearity: 5 runs at various intensities (ATWD 0, ATWD 1 separately). PSL FAT 3 Ped. Noise Time. Res Gain Lin. Opt. Sens #Runs@T 2 2 4 14 5 5 1 DOM MB 119 140 348 1062 298 226 bzip 2 MB 26. 7 40. 9 98. 5 297. 3 80. 1 38. 3 50 DOMs GB 6. 0 7. 0 17. 4 53. 1 14. 9 11. 3 bzip 2 GB 1. 33 2. 0 4. 9 14. 8 4. 0 1. 9 100 DOMs GB 12. 0 14. 0 34. 9 106. 3 29. 8 22. 7 bzip 2 GB 2. 7 4. 1 9. 8 29. 7 8. 0 3. 8 Total: 100 DOMs uncompressed 220 GB compressed 58. 1 GB Total: 5000 DOMs compressed 2. 9 TB M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 20

A 3: New Web Portal Face Easy-to-browse web Interface (java script). Existing content from earlier attempts transferred over. http: //uhe. phys. psu. edu/~testdaq/docs/ M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 21

A 4: Fat. Dark. Noise 0/5 Detailed results are extracted and presented on a webpage. See next pages for more plots M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 22

A 5: Fat. Dark. Noise 1/5 Main plot page. Temperature, scaler rates, arrival time distr. , duration, extended view arrival times. M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 23

A 6: Fat. Dark. Noise 2/5 Pulse heights after baseline level subtraction, leading edge times (ticks), average waveform, baseline level as found (ticks), RMS of average waveform, baseline difference between beginning and end of waveform M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 24

A 7: Fat. Dark. Noise 3/5 Baseline differences as a function of peak areas and of number of sub-peaks. M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 25

A 8: Fat. Dark. Noise 4/5 Four randomly selected waveforms displayed: Ch 0, 1, 2, 3, FADC (top to bottom) M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 26

A 9: Fat. Dark. Noise 5/5 Four randomly selected waveforms displayed: Ch 0, 1, 2, 3, FADC (top to bottom) As before, but here the Y-range is restricted and always the same. M Kestel, Collab. Maeeting: Dom. Test, Berkeley, Mar 20 27