Improved Parquet Reader
Summary
In this release we add optimized Parquet file importer. Previous versions read single Parquet file in a single importer process, in this version we improved it by virtually splitting files into fixed sized chunks that then can be imported in many parallel processes. For that, please do not forget to update to the latest deployment scripts in the user guide.
In addition, we added support for connections behind proxies when accessing cloud storage systems and added unified error codes.
Features
Refactoring
- #113: Added unified error codes
Dependency Updates
Compile Dependency Updates
- Added
com.exasol:error-reporting-java:0.4.0
- Updated
com.exasol:parquet-io-java:1.0.3
to1.1.0
- Updated
org.apache.orc:orc-core:1.6.9
to1.7.0
- Updated
org.alluxio:alluxio-core-client-hdfs:2.6.1
to2.6.2
- Updated
io.grpc:grpc-netty:1.39.0
to1.41.0
Test Dependency Updates
- Updated
com.dimafeng:testcontainers-scala-scalatest:0.39.5
to0.39.8
- Updated
com.exasol:exasol-testcontainers:4.0.0
to5.0.0
- Updated
org.mockito:mockito-core:3.11.2
to3.12.4
- Updated
org.scalatest:scalatest:3.2.9
to3.2.10
Plugin Updates
- Updated
com.eed3si9n:sbt-assembly:1.0.0
to1.1.0
- Updated
net.bzzt:sbt-reproducible-builds:0.28
to0.30
- Updated
org.scoverage:sbt-scoverage:1.8.2
to1.9.0