Skip to content


Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
As we have moved to Apache, please open your pull requests on:
Java Other

This branch is 7 commits ahead, 1065 commits behind master

Fetching latest commit…

Cannot retrieve the latest commit at this time

Failed to load latest commit information.

Parquet MR Build Status

Parquet-MR contains the java implementation of the Parquet format. Parquet is a columnar storage format for Hadoop; it provides efficient storage and encoding of data. Parquet uses the record shredding and assembly algorithm described in the Dremel paper to represent nested structures.

You can find some details about the format and intended use cases in our Hadoop Summit 2013 presentation


Parquet is a very active project, and new features are being added quickly; below is the state as of June 2013.

FeatureIn trunkIn devPlannedExpected release
Type-specific encodingYES1.0
Hive integrationYES (28)1.0
Pig integrationYES1.0
Cascading integrationYES1.0
Impala integrationYES (non-nested)1.0
Java Map/Reduce APIYES1.0
Native Avro supportYES1.0
Native Thrift supportYES1.0
Complex structure supportYES1.0
Future-proofed versioningYES1.0
Bit PackingYES1.0
Adaptive dictionary encodingYES1.0
Complex structure supportYES1.0
Predicate pushdownYES (68)2.0
Column statsYES2.0
Delta encodingYES2.0
Native Protocol Buffers supportYES2.0
Index pagesYES2.0

Map/Reduce integration

Input and Output formats. Note that to use an Input or Output format, you need to implement a WriteSupport or ReadSupport class, which will implement the conversion of your object to and from a Parquet schema.

We've implemented this for 2 popular data formats to provide a clean migration path as well:


Thrift integration is provided by the parquet-thrift sub-project. If you are using Thrift through Scala, you may be using Twitter's Scrooge. If that's the case, not to worry -- we took care of the Scrooge/Apache Thrift glue for you in the parquet-scrooge sub-project.


Avro conversion is implemented via the parquet-avro sub-project.

Create your own objects

  • The ParquetOutputFormat can be provided a WriteSupport to write your own objects to an event based RecordConsumer.
  • the ParquetInputFormat can be provided a ReadSupport to materialize your own objects by implementing a RecordMaterializer

See the APIs:

Apache Pig integration

A Loader and a Storer are provided to read and write Parquet files with Apache Pig

Storing data into Parquet in Pig is simple:

-- options you might want to fiddle with
SET 1048576 -- default. this is your min read/write unit.
SET parquet.block.size 524288000 -- your memory budget for buffering data
SET parquet.compression lzo -- or you can use none, gzip, snappy
STORE mydata into '/some/path' USING parquet.pig.ParquetStorer;

Reading in Pig is also simple:

mydata = LOAD '/some/path' USING parquet.pig.ParquetLoader();

If the data was stored using Pig, things will "just work". If the data was stored using another method, you will need to provide the Pig schema equivalent to the data you stored (you can also write the schema to the file footer while writing it -- but that's pretty advanced). We will provide a basic automatic schema conversion soon.

Hive integration

Hive integration is under development in Issue 28 and very close to being merged! Feel free to try out that branch and tell us what you think.


to run the unit tests: mvn test

to build the jars: mvn package

The build runs in Travis CI: Build Status

Add Parquet as a dependency in Maven

Snapshot releases


Official releases

We haven't published a 1.0.0 yet

Authors and contributors



Copyright 2012-2013 Twitter, Inc.

Licensed under the Apache License, Version 2.0:

Something went wrong with that request. Please try again.