Thing is there is a list of tables (350+ tables) on which I'm running the insert query i.e., INSERT INTO table ON CONFLICT DO NOTHING to continue inserting the records by ignoring the conflicting rows. But, at the same time I would like to capture the conflicting rows or every single conflicting column (not just the PK's) and their values for every single table.
INSERT INTO TABLE ON CONFLICT DO NOTHING don't even report us on the conflicting rows. How do I identify what the conflicting rows are by continuing the inserts to happen even if there are any conflicts. I mean my insert command shouldn't fail on conflicts but at the same time conflicts should be reported to the user.
Regards,
Nik
On 5/15/23 00:32, Nikhil Ingale wrote:
> Hi All,
>
> The following query inserts the rows by ignoring the rows that has conflicts.
>
> INSERT INTO test (id,name,age,branch) SELECT * FROM student ON CONFLICT DO
> NOTHING;
>
> How do I capture the conflicting records to a file while non conflicting
> records are inserted to the table?
On conflict insert the PK into a separate table, along with a timestamp
column populated by clock_timestamp(). (That way you can export and delete
sets of records while it's being written to.
--
Born in Arizona, moved to Babylonia.