Hi Tristan, you might want to include "--schemaprovider-class com.uber.hoodie.utilities.schema.FilebasedSchemaProvider" in the spark submit command. I also faced similar issue when I tried the Docker demo. I think there is a PR pending for Docs that includes this change.
Thanks, Sudha On Wed, May 1, 2019 at 1:33 PM Baker, Tristan <[email protected]> wrote: > Hi, > > Been working through the quickstart here: > https://hudi.apache.org/docker_demo.html > > I get an NPE when running the merge on read spark job. > > Here’s the spark-submit command (copied from the quickstart instructions) > > https://gist.github.com/tcbakes/4a11cff217fb8a98205b4cc46cd29750 > > > Here’s the NPE: > > https://gist.github.com/tcbakes/021258638184ddcbde2b0320ec589fde > > > I attached my debugger to the process and discovered that the > schemaProvider is null in on line 65 here: > > > https://github.com/apache/incubator-hudi/blob/3a0044216cb2f707639d48e2869f4ee6f25cfc19/hoodie-utilities/src/main/java/com/uber/hoodie/utilities/deltastreamer/SourceFormatAdapter.java#L65 > > The Copy On Write spark job/example works fine, but this one doesn’t. > > Any pointers? > > Thanks, > Tristan >
