Kuien Liu created HAWQ-1660:
-------------------------------
Summary: optimize parquet scan
Key: HAWQ-1660
URL: https://issues.apache.org/jira/browse/HAWQ-1660
Project: Apache HAWQ
Issue Type: Improvement
Components: Storage
Reporter: Kuien Liu
Assignee: Radar Lei
I saw Mr. Wen Lin's work on bloomfilter, and push down the fitler into parquet
scan. It provide a chance to accelerate further, that is: read the
hash-key-attributes first and perform filter, if positive, fetch the remaining
columns from parquet. does it make sense?
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
