问题
Background: I have an oracle table, the table doesn't have any specific column as timestamp while table creation script. This table can have millions of rows.
Example:
Employee {Emp_No, Name, Manager,Division, Role, Region}
My quest: If any updates happened through a job on that table, can i know which all rows got updated. Does oracle have any internal timestamp for each row which i can leverage. Can i use it in query to get all records.
Reason: I need to show my team those ambiguous records weren't updated by the job we all are suspecting.
回答1:
Oracle has ORA_ROWSCN Pseudocolumn. This columns returns "the conservative upper bound system change number (SCN)" of last transaction made on row or data block. This is a good estimate for when the block or row was last changed.
If your table is create with ROWDEPENDENCIES, ORA_ROWSCN returns scn for row. NOROWDEPENDENCIES is the default, in which case Oracle tracks SCN at the block level.
SCN_TO_TIMESTAMP allows you to converto scn to timestamp but for old scn it raises exception.
回答2:
Approximate update time can be retrieved with SCN_TO_TIMESTAMP(ORA_ROWSCN)
For each row, ORA_ROWSCN returns the conservative upper bound system change number (SCN) of the most recent change to the row. This pseudocolumn is useful for determining approximately when a row was last updated. It is not absolutely precise, because Oracle tracks SCNs by transaction committed for the block in which the row resides.
https://docs.oracle.com/cd/B19306_01/server.102/b14200/pseudocolumns007.htm
来源：https://stackoverflow.com/questions/43003024/get-all-tables-row-updated-after-a-specific-time