cdmikechen commented on issue #2005:
URL: https://github.com/apache/hudi/issues/2005#issuecomment-680351956
@bvaradar
Thanks for your reminder, I finally found my mistake:
I use hudi in a maven project with spark dependencies. I noticed that hudi
remove `com.twitter:parquet-hadoop-bundle`, so that I also removed this
dependency in my project.
```
<exclusions>
<exclusion>
<groupId>com.twitter</groupId>
<artifactId>parquet-hadoop-bundle</artifactId>
</exclusion>
</exclusions>
```
Therefore, when starting a spark task in this maven project, hudi can not
find `parquet-hadoop-bundle-1.6.0.jar` and `parquet.hadoop.ParquetInputFormat`
class. If I add a dependency, it should not report this error .
Meanwhile, I think my another suggestion which we should avoid new
`FileInputFormat` to just get class name should be fixed.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
