0

Consider the following:

  1. data are read from partitioned structure y=,m=,d=.
  2. hudi DataSourceWriteOptions.PARTITIONPATH_FIELD_OPT_KEY config is set to y=,m=,d=

At first glance I have to remove y,m,d columns but without these columns hudi will not know about partitions, no?

Cherry
  • 31,309
  • 66
  • 224
  • 364

2 Answers2

0

The answer is use version 0.9.0.

How to drop hoodie.datasource.write.partitionpath.field fields from a Hudi Dataset?

Hudi FAQ

Jira task with pull request link

Cherry
  • 31,309
  • 66
  • 224
  • 364
0

would you please open a github issue to get timely response.

sf lee
  • 1