UpcDst files were producing using upcDst code: https://github.com/adamjaro/star-upcDst
The output can be found here: /star/data01/pwg_tasks/upc02/UPC_P20ic

Although, less than 1k (out of 254k) muDst files were not processed in the upcDst production, many triggers have less than 90% of events compare to Jamie's page:
https://www.star.bnl.gov/protected/common/common2017/trigger2017/lumipp500GeV/lum_pertriggerid_pp2017_500GeV.txt

Hence, more muDst files were lost than expected!
There was parallel analysis of muDst files by Bill Schmidke.
Following tables compare number of events in upcDst files, at Jamie's webpage and number of events analyzed by Bill.

Trigger    #events     #events
ID         upcDst(M)  webpage(M)
570209      3.20        8.4     38.1%
570219      12.43       17.6    70.6%
570229      105.4       114.1   92.4%
570701      182.9       233.0   78.5%
570702      7.61        10.0    76.1%
570703      25.81       31.5    81.9%
570704      9.30        10.0    93.0%
570705      628.1       670.8   93.6%
570709      85.99       99.1    86.8%
570711      36.65       39.8    92.1%
570712      34.92       39.9    87.5%
570719      84.30       91.3    92.3%

Comparison with Bill's numbers in M
Trigger         upcDst   Bill's   Jamie's
JPsi*HTTP     121    130    140
RP_UPC         42.5         45      50
ZB                 9.3       9.8     10