Hi,
just to elaborate what Ranadip has pointed out here correctly, gzip files
are read only by one executor, where as a bzip file can be read by multiple
executors therefore their reading speed will be parallelised and higher.
try to use bzip2 for kafka connect.
Regards,
Gourav Sengupta
On Mon,
Dear all,
As you may know we at JetBrains develop Kotlin API for Apache Spark [1].
It's stable for some time already, version 1.0 was released more than a
year ago. Also we've released version 1.1 [2] with support for Spark
Streaming, RDDs and Jupyter several days ago.
We believe that there are
Unsubscribe
