Processing an S-Q split

 

Pre-conditions

Initial podfile received and processed, leaving

 

2nd podfile received and processed, leaving

 

Solution

  1. Identify the 2nd podfile by looking up the exposure name of the DP=f exposure in DSQUERY::OPUS_DB:podnames.

  2. Run the podwhacker tool on 2nd podfile to identify all exposures (there can be several) that are coming in this 2nd podfile AND that have corresponding DP=p OSFs from the 1st podfile.

    % podwhacker podfilename |& awk ‘{print $2}’ | sort | uniq

     

  3. Use merge_reqd.pl tool to load the names of all exposures found from the previous step into the otfr_special_processing table.

     

  4. Use the OMG to modify the S-version OSF for each exposure from step#2. Change DP=e, so that automatic Q-S split merging does not occur prematurely for these exposures.

     

  5. If duplicate checking is on for podnames (PODNAMES_CHECK = TRUE), make sure OPUS_DB's podnames relation does not contain this exposure. Delete it if it does. Otherwise DP will go to d, for duplicate,.

     

  6. Resubmit the 2nd podfile for processing by DP (rename/move the pod file for pipeline pick-up). DP should

    • Allow the Q-S precedence rule to be bypassed for each exposure entered in otfr_special_processing (a message shows in the process logfiles)
    • One or more exposures (Q-versions) with DP=p
    • One or more exposures (S-versions) with DP=e (the ones changed in the previous step)
    • Incomplete EDT sets for both the Q and S versions in the SIS_DIR

     

  7. Use the OMG to modify the S-version OSFs from step#3 from DP=e to DP=p. They should now get automatically detected for Q-S split merging by the FNDMRG process, and get merged by the MRGxxx process.

 

Clean-up

OSFs for Q-S split exposures that were not sent down the pipeline (those with the smaller DCFNUM field) remain DP=c, with no further processing. These OSFs should be deleted.