How to restart Delta Loads for GL, AR, AP and COPA R/3 extracts:
When the failed load is restarted, it will go for ‘repeat request of last delta’ will bring Zero records. The next schedule will be a Delta. So the recovery process would be first schedule a ‘Repeat request of Delta’ from the info package, next schedule a ‘Delta’ by doing a repeat on the process chain OR by Restarting the Maestro job
Need to change Technical Status for the request in failed status to red before you RESTART the failed DELTA load.
Before deleting failed request, prevent further problems by always manually resetting the technical status for the data request to Red. This ensures that BW properly communicates back to source that the request has failed. Without performing this step, you may not see Repeat as an option on the failed info package in the process chain or a repeat of the failed info package may not return records the first time it is re-executed. Some times you have to refresh the process chain for the Repeat step to appear after you’ve changed the technical status.
Select the Failed request in the Info cube/DSO, Hit Delete and select Yes when prompted.
It will be always safe to check the timestamp entry in ECC before you restart the failed DELTA load.
How to check the timestamp in ECC:
Go to the transaction SE16, display the table BWOM2_TIMEST:
Give the data source name ,for example0FI_GL_4:
The field “Last Timestamp-LAST_TS” will be marked with ‘X’,this entry corresponds to the last successful run of either Delta extract or BW extract.
The Delta extract runs can be checked in SM37 ,job name is ‘BIWDCRFIGL_DELTA_EXTRACT’ or Maestro schedule for job stream ‘BIR3FIGLDELTADLY ‘.Compare the the runs with the entries in table BWOM2_TIMEST.Also check the last successful BW extract and compare the entry with timestamp table. If both the entries are matched which determines the timestamp table got updated as expected and didn’t loose any delta.
Since the request is already made red and deleted from the Pass Thru, restart the load.
This scenario, some times the ‘repeat request of last delta’ will bring Zero records. The next schedule will be a Delta. So the recovery process would be first schedule a ‘Repeat request of Delta’ from the info package, next schedule a ‘Delta’ by doing a repeat on the process chain.