Production status and outstanding questions Ivan Furic, Ken Herner 31 Aug 2018
Keepup pipeline • Updating DC2-style processing to work on new raw data (w/compression) in dunetpc v07_0[3-4]_00 – All decoder modules and caldata reco appear to work fine – Nothing in the reco chain after that seems to work so far (that I have tried anyway) – Q1: start this way, or try to add more reco modules? • Everything in the same file leads to more than doubling of the size (a 5 GB raw file outputs a 13 GB file without dropping the raw; that was about 400 events) – Q2: drop the raw from the output files? – Q3: Should we do any other data integrity checks within these jobs? 2 8/31/18 K. Herner | Production Meeting
Keepup pipeline (2) • Decoder output will presumably go to a separate file family (easy deletion if so desired) • Q4: need a dropbox path and file family decided • Q5: What kind of datasets will be useful for people looking at this data? • Submitted request today to double DUNE Fermigrid quota for 3 months. Need to assign quota fractions to the various subgroups – Propose production gets at least 90% of overall quota – Prod.keepup should have priority over prod.mcsim. Maybe 80/20? 90/10? 3 8/31/18 K. Herner | DUNE Production
MCC11 Update • Finished atmospheric samples; only protoDUNE samples left • New requests from proton decay group this week – Still figuring out exactly what they want and what resources the jobs require – Unclear if this will finish before beam • Still did not hear from main physics groups on samples • Proposal: Declare requests submitted after 9/7 to be best- effort only until protoDUNE run finished (meaning they will have reduced priority). We can run them in off-weeks when no outstanding data to process. 4 8/31/18 Presenter | Presentation Title or Meeting Title
Recommend
More recommend
