[jira] [Updated] (HUDI-4967) Improve docs for meta sync with TimestampBasedKeyGenerator

["Ethan Guo (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HUDI-4967?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Ethan Guo updated HUDI-4967:
----------------------------
    Fix Version/s: 0.14.0
                       (was: 0.13.0)

> Improve docs for meta sync with TimestampBasedKeyGenerator
> ----------------------------------------------------------
>
>                 Key: HUDI-4967
>                 URL: https://issues.apache.org/jira/browse/HUDI-4967
>             Project: Apache Hudi
>          Issue Type: Improvement
>            Reporter: Ethan Guo
>            Assignee: Jonathan Vexler
>            Priority: Major
>              Labels: pull-request-available
>             Fix For: 0.14.0
>
>
> Related fix: HUDI-4966
> We need to add docs on how to properly set the meta sync configuration, especially the hoodie.datasource.hive_sync.partition_value_extractor, in [https://hudi.apache.org/docs/key_generation] (for different Hudi versions, the config can be different).  Check the ticket above and PR description of [https://github.com/apache/hudi/pull/6851] for more details.
> We should also add the migration setup on the key generation page as well: [https://hudi.apache.org/releases/release-0.12.0/#configuration-updates]
>  * {{{}hoodie.datasource.hive_sync.partition_value_extractor{}}}: This config is used to extract and transform partition value during Hive sync. Its default value has been changed from {{SlashEncodedDayPartitionValueExtractor}} to {{{}MultiPartKeysValueExtractor{}}}. If you relied on the previous default value (i.e., have not set it explicitly), you are required to set the config to {{{}org.apache.hudi.hive.SlashEncodedDayPartitionValueExtractor{}}}. From this release, if this config is not set and Hive sync is enabled, then partition value extractor class will be *automatically inferred* on the basis of number of partition fields and whether or not hive style partitioning is enabled.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)