and you get list of build paths:
/Volumes/DRS_Staging/DRS/prod/20180716/buildList1.txt.23.33/batchW00EGS1015752-1n /Volumes/DRS_Staging/DRS/prod/20180716/buildList1.txt.23.33/batchW00KG0624-1n /Volumes/DRS_Staging/DRS/prod/20180716/buildList2.txt.23.33/batchW1KG14500-1n /Volumes/DRS_Staging/DRS/prod/20180716/buildList2.txt.23.33/batchW1KG14783-1n
Now look for error summaries which do not have sequence errors ( the most common kind)
while read ee ; do grep -i seQUEnce $ee-logs/errorSummary.txt; (($? == 1 )) && { echo $ee >> fixable.lst;} ; done < lst
Now fixable.lst has only the batches we can rebuild in place.
This won't work. problem is you can't restart an arbitrary batch, because project.conf has been written over between batches. Log it as a fail and try again.
Sometimes batch builds fail because of a resource error. Rather than rebuild and recopy, you can just retry after the copies. Find them first
find $PR/20180717 -name errorSummary.txt -maxdepth 3 -not -size 0 -ls > ~/tmp/20180717errors/lst
awk -F' ' '{ print $NF }' ~/tmp/20180717errors/lst | sed 's/-logs.*//'`
and you get list of build paths:
/Volumes/DRS_Staging/DRS/prod/20180716/buildList1.txt.23.33/batchW00EGS1015752-1n /Volumes/DRS_Staging/DRS/prod/20180716/buildList1.txt.23.33/batchW00KG0624-1n /Volumes/DRS_Staging/DRS/prod/20180716/buildList2.txt.23.33/batchW1KG14500-1n /Volumes/DRS_Staging/DRS/prod/20180716/buildList2.txt.23.33/batchW1KG14783-1n
Now look for error summaries which do not have sequence errors ( the most common kind)while read ee ; do grep -i seQUEnce $ee-logs/errorSummary.txt; (($? == 1 )) && { echo $ee >> fixable.lst;} ; done < lst
Now fixable.lst has only the batches we can rebuild in place.Run fix-one-batch against it.