19 Sep 2016
Fix: Dropping a partition while inserting by primary key into same partition should not raise an exception but returns a 0 row count instead.
Fix: JAR conflict if HDFS plugin with Hadoop 2.x is used.
Fix: Now it is also possible to restore a snapshot of a single partition of a partitioned table in a non-existent table. This requires that the snapshot has been created with version
0.55.5 or greater!
Fix: Join conditions were not properly validated and an expression with a not yet defined relation could be used.
Fix: Setting of a default schema was not possible by using the HTTP REST endpoint.
Fix: Error was thrown in case of primary key bulk delete statement with null bulk arguments.
Fix: Return 0 rows instead of an error when executing select queries that try to match null for primary key columns. eg: select * from t where pk in (null);
Fix: Allow select statements on columns of multiple nested object array
Fix: do not allow to create tables in read-only system schemas
COPY FROM statements fail if null values are inserted into partitioned tables.
Fix: type was not set correctly when creating a token filter with default name and no type
Fix: prevent creating
char filters with missing properties