We have a scenario in production where in a data load from DSO A to DSO B is taking 2H 52M, out of which 2H 48M is being taken by "Start of extration" process for Data Packet 1 itself. Whereas for remaining data packets it runs within 2mins. We checked the SQL trace, Jobs in SM37, rsrv, statistics but could not find any resolution. Number of records in DSO A (Source) is 683,423. I went through below link as well, but it wasnt of any help.
Can please someone help to identify a way to optimize this data load.