During the investigation of ldbc/ldbc_snb_datagen_hadoop#24, we found anoter potential issue. Namely, if the TCR of the benchmark is set to e.g. 0.001, the delta time (10 seconds) used as a safety margin between updates, can shrink to as low as 10 milliseconds. This may not be enough for the update to finish and it can cause systems to abort transactions.
Therefore, we should consider increasing the delta time.