skip bad record in redshift data load

roy picture roy · May 12, 2014 · Viewed 16.1k times · Source

I am trying to load data into AWS redshift using following command

copy venue from 's3://mybucket/venue'
credentials 'aws_access_key_id=<access-key-id>;aws_secret_access_key=<secret-access-key>'
delimiter '\t';

but data load is failing, when I checked Query section for that specific load I noticed it failed because of "Bad UTF8 hex sequence: a4 (error 3)"

Is there a way to skip bad records in data load into redshift?

Answer

mike_pdb picture mike_pdb · May 13, 2014

Yes, you can use the maxerror parameter. This example will allow up to 250 bad records to be skipped (the errors are written to stl_load_errors):

copy venue from 's3://mybucket/venue' credentials 'aws_access_key_id=;aws_secret_access_key=' delimiter '\t' maxerror as 250;