Check availability of tape-lost files at CNAF

[mt82]

@Matteo Tenti I just added you to the watch list of INC000001183691, the ticket has as attachment two list of files that we would like to check if they are available at CNAF and possibly restore them at FNAL too

[mt82]

FB1699L9.files.NOEXISTING FB1699L9.files.NOPOOLs Both have to be checked

[mt82]

FB1699L9.files.NOEXISTING.CNAF.not.existing.txt FB1699L9.files.NOEXISTING.CNAF.existing.txt FB1699L9.files.NOPOOLs.CNAF.not.existing.txt FB1699L9.files.NOPOOLs.CNAF.existing.txt

[mt82]

Proposal:

1. create a RUCIO dataset including all the missing files
2. add a rule for a replica at FNAL_DCACHE RSE
3. move files from FNAL_DCACHE to the final location

[mt82]

• recall files from tape at CNAF
• develop a script:
  • filename as input
  • get uri from rucio at cnaf
  • get original location at fnal from samweb
  • construct ifdh cp command

[mt82]

@aheggest

• im taking a look right now. it seems that the first run in your list run 9885 has a comment in the Trigger/Beam runs spreadsheet “9877-9885 are PMT debugging efforts after server self-reboot and subsequent causality warning clearing” (edited)

• i just compared your run list with the runs spreadsheet and it seems all of the runs in the list you shared were marked “N” for keepup or the run seemed to crash/not all components were included. here are the comments for the few runs in your list that were marked “Y” for keepup:

  • run 11230 was marked “Y” for keepup, but it crashed “Trigger test run with NuMI early warning signals. Crashed due to lost process icaruspmtewbot01"
  • run 11231 was marked “Y” for keepup, but it stopped “Trigger test run with NuMI early warning signals. Run stopped as NuMI signals no longer provided.”
  • run 11291 was marked “Y” for keepup, with comment “Trigger test. PMT only”
  • run 11573 was marked “Y” for keepup with comment “Calibration run without EE[13-19]”
  • for runs 9901, 9909 and 9910 , i see a comment “runs 9898-9910 are trigger test runs with no beam active”

  so it seems to me there are no good runs in the list you shared