Mails connected with RoIB test runs ----------------------------------- Edited mail from Per Werner Date: 29 September 2004 11:15:10 BST Subject: RoIBuilder with L1Calo and L1Muon trigger input This morning we wrote two files with both trigger sources (run 1031 and 1034) and two with only the calo trigger input (1032 and 1033). Due to problems with BCID checks in the Muon trigger, ROS-101 was not (always) providing data => files for run 1031 and 1034 were not closed (have extension 'writing') as the SFI crashed and can't be read with 'readData'. [NB Szymon told us changing extension from 'writing' to 'data' fixes this]. We should put a low limit on the no.of events in the L2SV to avoid crashing SFI and getting a proper data file. The tar file tdaq_logs.RoiBTest_L1Calo_MuCTPI.tgz contain the directory 'tdaq_logs' which contain the four data files daq*103[1-4]* and all the output from the Online and Dataflow SW. Cheers, Per ================================================================================= Exchange of mails about analysis of RoIB test runs -------------------------------------------------- From: Murrough Landon Date: 19 October 2004 14:27:03 BST Hi Per, Did you or Jim look at the data we took in that test with the RoIBuilder? Im afraid I havent. A few L1Calo people are having a phone conference to talk about progress with analysing the data we took in the combined runs during the 25ns period. If there were some conclusions from the RoIB test run data, I could report it to my colleagues. Cheers, Murrough. From: Per Werner Date: 19 October 2004 17:08:57 BST Hello Murrough, I didn't but Bob Blair did. There was no Muon data but I assume that the L1Calo was there. Mayby Bob can expand. Cheers, Per From: "Robert E. Blair" Date: 19 October 2004 20:51:14 BST There is L1Calo data, but I did not look too hard to try to figure out whether it is "sensible". I simply used Thorsten's perlscript to dump the data. I'm happy to look a little deeper if I get some guidance on what to look for. From: Murrough Landon Date: 19 October 2004 21:17:28 BST Hi Bob, Thanks for the info. As I recall we sent very simple constant RoIs every event. So as long as every event seen in the RoIB contained an L1Calo fragment with no errors, timeouts etc, thats about all we can usefully check - though perhaps you could check that in fact the RoI was the same every time. I dont have our logbook (its at Rutherford) so I cant be sure exactly what the RoI fragment should contain but I think there should be just one RoI word per fragment. Unfortunately there wasnt enough time during the testbeam period to do more stressful tests, eg with varying data, higher rates, busy etc. Cheers, Murrough.