New Interfaces in Hive 2.1

Hive 2.1 includes the following new interfaces:

Interface Description
org.apache.hadoop.hive.common.Pool Simple object pool to prevent GC on small objects passed between threads. An allocator provided externally to storage classes to allocate MemoryBuffer-s. An abstract data cache that IO formats can use to retrieve and cache data.
org.apache.hadoop.hive.conf.HiveVariableSource Getting hive variables.
org.apache.hadoop.hive.metastore.FileFormatProxy Same as PartitionExpressionProxy, but for file format specific methods for metadata cache.
org.apache.hadoop.hive.metastore.HouseKeeperService Runs arbitrary background logic inside the metastore service.
org.apache.hadoop.hive.metastore.txn.TxnStore A handler to answer transaction related calls that come into the metastore server.
org.apache.hadoop.hive.ql.exec.tez.KeyValuesAdapter Key-values interface for the Reader used by ReduceRecordSource
org.apache.hadoop.hive.ql.exec.vector.expressions.IStructInExpr Interface used for both filter and non-filter versions of IN to simplify VectorizationContext code.
org.apache.hadoop.hive.ql.exec.vector.expressions.ITimestampInExpr Interface used to proccess timestamp in expression. Interface when implemented should return the estimated size of columnar projections that will be read from the split. Split that is aware that it could be executed in LLAP. Marker interface for LLAP; serves no other purpose. Marker interface to indicate a given input format is self-describing and can perform schema evolution itself. Marker interface for streaming output formats. This interface has functions that provide the ability to customize the translation from Hive internal representations of Authorization objects to the public API objects This is an interface that is not meant for general use, it is targeted to some specific use cases of Apache Sentry (incubating).
org.apache.hadoop.hive.serde2.FetchFormatter (For internal-use only) Used in ListSinkOperator for formatting final output.