----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/70600/#review215207 -----------------------------------------------------------
Hi Jungtaek, I spoke to Ankur on Friday and he mentions all spark tables/db/cols in CDP will always be created in HMS. If that is the case we should remove spark tables/db/cols models. In case spark will use Hbase to create thsee then we should map them to hbase models that we have in Atlas. Ankur could you please comment on the same. Thanks - Aadarsh Jajodia On May 10, 2019, 9:18 p.m., Aadarsh Jajodia wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/70600/ > ----------------------------------------------------------- > > (Updated May 10, 2019, 9:18 p.m.) > > > Review request for atlas, Ankur Gupta, Ashutosh Mestry, Jungtaek Lim, Sridhar > K, Le Ma, Madhan Neethiraj, and Sarath Subramanian. > > > Repository: atlas > > > Description > ------- > > This is an initial mode file for the SPARK models. Please refer to the file > added here. This file is a GET call on typeDefs on a cluster which had SAC > enabled and SAC has created the models. The SAC version used was > https://github.com/hortonworks/spark-atlas-connector/tree/SAC-CDP-1.0-dev > > > Diffs > ----- > > addons/models/1000-Hadoop/1090-spark_model.json PRE-CREATION > > > Diff: https://reviews.apache.org/r/70600/diff/1/ > > > Testing > ------- > > > File Attachments > ---------------- > > SAC Models > > https://reviews.apache.org/media/uploaded/files/2019/05/07/e31827b2-04c5-4fab-b5b7-4271398831e4__spark_sac_api_model_definition.txt > > > Thanks, > > Aadarsh Jajodia > >
