Hello all,

We have the multi-engine support PR up here
https://github.com/apache/hudi/pull/1827/files
<https://github.com/apache/hudi/pull/1827/files#diff-ea38810925fa8d48dfec1715bc182c32>

While I plan to spend good amount of time reviewing the same, I also
encourage everyone to take a look and suggest changes.  This a monumental
and critical change for the project, given it changes some of the most
fundamental abstractions since Hudi's creation.

Specifically, folks at Uber/Yotpo (who are the RDD API users AFAIK), please
review the changes here to HoodieWriteClient and the now
HoodieSparkWriteClient

https://github.com/apache/hudi/pull/1827/files#diff-ea38810925fa8d48dfec1715bc182c32

leesf, liwei and mathieu (authors) - can you test the spark path more
(datasource and deltastreamer) and ensure existing functionality works as
expected..

Any other ideas welcome to land this safely and set us up for a Flinky
future.

thanks
vinoth

Reply via email to