Zeyu Wang created HUDI-7642:
-------------------------------
Summary: Compact MOR tables with operation fields cause data errors
Key: HUDI-7642
URL: https://issues.apache.org/jira/browse/HUDI-7642
Project: Apache Hudi
Issue Type: Bug
Reporter: Zeyu Wang
When we compact an MOR table who with _hoodie_operation field, the hoodiekey
tagged with operation "-D" was not correctly removed.
Refer to previous discussions
(https://github.com/apache/hudi/pull/8721#issuecomment-1736629662) we should
keep flink engine for the delete record, And also repair the spark in the
https://github.com/apache/hudi/pull/10219 engine problems when reading data,
should repair caused by compact problem now. Because of the 'compact' directly
using the HoodieMergedLogRecordScanner in the common module, I think we have
to add some optional configuration to control whether or not the
HoodieMergedLogRecordScanner directly delete the key that taged with "-D"
operation
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
