Thanks for OpenInx's answer. How about progress of integration with Flink? I thought that Flink have many users in China, but lack of data lake. This integration gives more benefit for Flink users.
At 2020-01-07 17:05:48, "OpenInx" <[email protected]> wrote: >Hi Nicholas > >Integrating the flink with hudi is the on-going work what we are doing, >especially VinoYoung is pushing the thing >forward, started the discussion in JIRA [1], and provided a draft document >in here [2]. > >[1]. https://issues.apache.org/jira/browse/HUDI-184 >[2]. >https://docs.google.com/document/d/1wDTpzuehfcDYMTD3c-lJqIiP1rt095XBzdZoPYMTvlY/edit > >On Tue, Jan 7, 2020 at 4:37 PM 蒋晓峰 <[email protected]> wrote: > >> Hi Syed Abdul Kather, >> About supporting flink, is it necessary to abstract api for integration >> with Spark and Flink? Currently Hudi is strongly bind with Spark, not >> pluggable. To support Flink needs abstract API adapter to any engine. >> >> >> Thanks, >> Nicholas >> >> At 2020-01-07 16:26:13, "Syed Abdul Kather" <[email protected]> wrote: >> >No, complete the hudi runs on spark. There is a plan to support flink but >> >not sure. >> > Thanks and Regards, >> > S SYED ABDUL KATHER >> > >> > >> > >> >On Tue, Jan 7, 2020 at 12:59 PM [email protected] <[email protected]> >> wrote: >> > >> >> Hi, >> >> Can Hudi insert / upsert without Spark? >> >> >> >> >> >> >> >> [email protected] >> >> >>
