DRC CHANGEOVER DEC TRISTAN TINGJUN T H 30

DRC CHANGEOVER DEC TRISTAN TINGJUN T H 30

2 Trigger config - rces_and_ssps_iupulse(_nodiskwrite) Applications config - docs/config. txt Recovered from run 6293 – the overnight run which lasted 6 hours with nodiskwriting Ran ‘disk write’ and ‘no disk write’ runs all day with rces 08 -15 and ssps 01 -07 Yujing wrote to disk for runs 6309 -6322. 12/14 stayed in sync. See shift summary 3270 in ECL. Shannon used the same run settings but did some runs omitting RCE 10 which was giving errors. By end of shift RCE 10 was working again. Run 6376 left overnight DEC 26 TH

3 Trigger config - rces_and_ssps_iupulse(_nodiskwrite) Applications config - docs/config. txt Recovered from run 6376 – the overnight run which lasted 5 hours with nodiskwriting. Split day/evening shift into becnhmarking RCE’s 08 -11 and 12 -15 respectively. Ran ‘disk write’ and ‘no disk write runs’ during the day shift with rces 08 -11 and ssps 01 -07. Some runs with all RCES enabled. See ECL entry 3297. Ran ‘disk write’ and ‘no disk write runs’ during the evening shift with rces 12 -15 and ssps 01 -07. Some runs with all RCES enabled. See ECL entry 3310. Nuno did some PTB tests at 11: 39 – ECL entry 3290 Yujing wrote to disk for runs 6309 -6322. 12/14 stayed in sync. See shift summary 3270 in ECL. Shannon used the same run settings but did some runs omitting RCE 10 which was giving errors. By end of shift it was working again Run 6410 left overnight DEC 27 TH

4 Trigger config - rces_and_ssps_iupulse(_nodiskwrite) Applications config - docs/config. txt Recovered from run 6410 – the overnight run which lasted 4 hours 44 mins with nodiskwriting. Only Yichen on shift, only one shift. Ran ‘disk write’ with rces 08 -15 and ssps 01 -07. See ECL entry 3297. Changed to ‘no disk write’ runs as of run 6424 due to continuous crashes, changed back to disk writing as of run 6427. Changed back to no disk writing as of run 6437. Michelle changed which CCU’s were powered at 10 am – See ECL 3317 Tristan took over shift at 3 pm to allow Yichen to start uboone shift – did two no disk write runs. Run 6450 left overnight Kurt did some throughput tests at 4 pm – See ECL entry 3337 DEC 28 TH

DEC 29 TH Kurt Entry 3342 “To reduce problems with RCE board readers running out of buffers, I've made two changes: * added 6 more Event. Builders in config. txt (total is now 8) * changed the RCE raw_buffer_precommit value from 100 to 2000 These values are based on throughput studies that we have done recently, and we may refine them as we learn more. I believe that these changes would definitely helped avoid the crash that happened at 17: 52 last evening” Trigger config – not constant, work began on PTB after go ahead from Nuno. Applications config – not constant, as above. Started day with rces_and_ssps_iupulse daq config and normal docs/config. txt. Then, later, used the same daq config but with the docs/config_newyear. txt. This config is desgined to test the latest merge of Nunos PTB software with the latest changes to ‘develop’. Ran ‘disk write’ with rces 08 -15 and ssps 01 -07 with the newyear config file. Had the most stable runs as a result – didn’t crash, actually stopped the DAQ. Excellent! Stopped at run 6490 so I could go to PC 4 and test some PTB stuff at 13: 24. PTB tests were unsuccessful. I managed to get one count of PTB triggers with CCU 1 turned on. This was 363, or 36. 3 Hz. I recall two unrecorded counts both at ~300 counts in 10 s. Due to board reader errors, stopped work on this at 17: 45. Managed to take one run with penn 01 (PTB) enabled and rces and ssps off. This uses the config_newyear. txt application config and the rces_and_ssps_and_ptb daq config. This was run 9497. Managed to take one run at 22: 48 that uses the same settings as above, with the rces and ssps turned on as well. The PTB did not give output during this time (board reader errors) but the DAQ did actually run with all the components enabled. Half success. Run # 6500. Waiting on Nuno to arrive to debug board reader errors Only got errors when trying to start runs like 6497 (PTB only) and like 6500 (PTB & RCEs and SSPs). 5

6 Tingjun took over as DRC. At around 11 am I went to PC 4 and switched back to the random trigger from the PTB trigger in order to take working runs again. Will wait on Nuno to arrive to proceed with more PTB work. Random trigger gave an average (over 10, 10 s samples) of 5. 8 Hz. Changed counter overlap to give an average trigger rate of 17. 6 Hz (over 10, 10 s samples). Now iupulse runs can be used to benchmark failure of disk/no-disk writing today. Increased rate does not appear to increase how much data can be taken before a crash. But data is taken at a faster rate, so DAQ crashes in less time. More careful investigation needed. DEC 30 TH