> https://issues.apache.org/jira/browse/PARQUET-2193 .
> Basically, a workaround exists today, please check if it works for
> you.
Indeed, setting the bellow code in the reader side is a working
workaround. Thanks for looking for more general solution.
```
spark.sparkContext.hadoopConfiguration.set("parquet.split.files",
"false")
```
> (in the future, please send emails with text, instead of attaching it
> as afile).
sorry for this, I tried a new feature of my email client (evolution)
based on markdown, and this is attaching a html file, my bad.
