The last 2 weeks I’ve had two deal with two instances where the daily ii jv feed didn’t balance. Usually I get a call from controllers asking me to investigate. Like almost always… both instances were due to an error in the original spreadsheet that was loaded into Banner. For these two cases the datestamp for the particular VAxxxxxx jv was not the same across all seqno records. These records all made it into the gurfeed table. The first time this happened, fwpfedh.sql did not put header records into gurfeed. These are the record type 1 records. There should be one of these for each jv. The second time this happened, fwpfedh.sql put header records into gurfeed but the total only went in for one of the timestamps. So when FURFEED runs on the institution side it can’t handle it.
The first case left all the records in gurfeed and didn’t process anything. To fix I added a header record manually with a correct total and then ran furfeed, fgrtrni, fgrtrnr,fgbactg in the institution instance. I had to get the big schools to do it since I don’t have access to their databases. The second case hit every institution and fed only the part of the jv that matched the header timestamp. Since this posted, t required a new correcting jv referencing the old one to post the missing rows. We deleted the orphan records in gurfeed.
This was a lot of work for everybody so I made a mod to iitfeed.pc to throw out any jvs that did not have a matching timestamp in all the seqno rows for a particular jv. The jv should never load into owag to be split up.
