Alexey Kudinkin created HUDI-2788:
-------------------------------------
Summary: Z-ordering Layout Optimization Strategy fails w/ Data
Skipping enabled
Key: HUDI-2788
URL: https://issues.apache.org/jira/browse/HUDI-2788
Project: Apache Hudi
Issue Type: Bug
Components: Index
Reporter: Alexey Kudinkin
Fix For: 0.10.0
During testing of Z-ordering in test environment i've discovered following
issues:
# Queries failing for tables w/ enabled Clustering w/ Z-ordering Layout and
data-skipping enabled, being unable to read `_SUCCESS` file (automatically
created by Spark)
# Some of the translations of the original query predicates into Z-index table
predicates are translated incorrectly (`!=`, `not like`, etc)
# Join merging indexes across commits incorrectly always checks for null first
column (instead of Nth column) when picking the result of the merge
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
