[IMPALA-3145] Refactor the IGNORE clause

• Type: Improvement
• Status: Resolved
• Priority: Minor
• Resolution: Duplicate
• Affects Version/s: Kudu_Impala
• Component/s: Frontend
• Labels: kudu usability
• Target Version: Kudu_Impala
• Assignee: Dimitris Tsirogiannis
• Reporter: David Alves
• Created: 04/Mar/16 21:20
• Updated: 19/Oct/16 17:22
• Resolved: 19/Oct/16 17:22

Description

IGNORE is being used to refer to ignoring errors when inserting keys that already exist or updating rows that don't exist yet.

Marcel points out that, in some circumstances, ignoring should be the default.

Moreover:
Marcel:
my point wasn't purely that 'ignore' should be the default, but simply saying 'ignore' doesn't point out what is being ignored. i think this should be 'ignore duplicates'.