Repeated benchmark runs would use the same ID as previous runs, which the backend would ignore. Update the used IDs to fix this and also distinct it from ordinary, non-benchmark program runs..

Submitted by Elmer van Chastelet on 12 May 2023 at 13:40