[
https://issues.apache.org/jira/browse/HUDI-6873?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jonathan Vexler updated HUDI-6873:
----------------------------------
Description: If the payload is overwritewithlatestavropayload this matters
because if the base file and the update have the same precombine, then the
record in the base file will be used instead of records from later writes
(was: If new columns are added using OOB schema evolution on an MOR table and
then clustering is performed, the field will be null for all records after
clustering. This has only been tested with deltastreamer, but may also affect
other write sources as well.)
Priority: Critical (was: Major)
Summary: Clustering MOR applies base files after log files (was:
Clustering MOR with OOB schema evolution results in data loss)
> Clustering MOR applies base files after log files
> -------------------------------------------------
>
> Key: HUDI-6873
> URL: https://issues.apache.org/jira/browse/HUDI-6873
> Project: Apache Hudi
> Issue Type: Bug
> Components: deltastreamer, spark
> Reporter: Jonathan Vexler
> Assignee: Jonathan Vexler
> Priority: Critical
>
> If the payload is overwritewithlatestavropayload this matters because if the
> base file and the update have the same precombine, then the record in the
> base file will be used instead of records from later writes
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
