echauchot commented on PR #3:
URL: 
https://github.com/apache/flink-connector-cassandra/pull/3#issuecomment-1397142509

   > > I did not author the RingRange and SplitsGenerator classes. I got them 
from the Apache Beam Cassandra connector.
   > 
   > If you're copying stuff from other projects it please add some 
notice/attribution to the files and update the Flink source notice accordingly.
   
   ah yes I forgot this Flink guideline. I don't think it is an ASF rule but 
fair enough, I'll add it to the javadoc and notice.
   
   > 
   > > Back in 2017 I coded a [splitter for Cassandra Beam 
connector](https://github.com/echauchot/beam/blob/bfa33b85b6b310556ffa5c44c99bef50575b2c56/sdks/java/io/cassandra/src/main/java/org/apache/beam/sdk/io/cassandra/CassandraIO.java#L346)
 that works [with 
tokens](https://github.com/echauchot/beam/blob/BEAM-245-CassandraIO/sdks/java/io/cassandra/src/main/java/org/apache/beam/sdk/io/cassandra/DataSizeEstimates.java)
 also but that is simpler and supports all the Cassandra partitionners. Would 
you prefer that we use this other approach ?
   > 
   > Not sure? Why didn't it make it into Beam? Do you know why the Beam code 
is written the way it is?
   
   Actually, another splitting approach was opted in by the reviewer in 2017. 
But short after there was another author who changed the splitting to something 
similar to my 2017 token based code. So when I thought about coding the split 
for Flink connector I decided to take the version of the code that was merged 
to Beam master. But it is true that it is over complicated, redundant and not 
supporting the non-default Cassandra partitioner. The approach I had in 2017 
was the same as the Cassandra Spark connector written by datastax (tokens + 
cassandra size estimates statistics). So I think I'll try to reuse this code, 
migrate it to Flink and update it to the latest Cassandra version and push it 
in this PR. 
   WDYT ?
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to