Name Apache Hudi
Apache Parquet
Description Apache Hudi is a transactional data lake platform that brings database and data warehouse capabilities to the data lake. Utilises data stored in either parquet or orc.
Apache Parquet is an open source, column-oriented data file format designed for efficient data storage and retrieval.
Source code https://github.com/apache/hudi https://github.com/apache/parquet-format
Website https://hudi.apache.org/ https://parquet.apache.org/
License Apache license 2.0
Apache license 2.0
Year created 2016
2013
Company Uber
Twitter, Cloudera
Use cases Incremental data processing, Data upserts, Change Data Capture (CDC), ACID transactions
Write once read many, Analytics, Efficient storage, Column based queries
Language support java, scala, c++, python, r, php
Is human readable no
no
Orientation column or row column
Has type system yes
yes
Has nested structure support yes
yes
Has native compression yes
yes
Has encoding support yes
yes
Has constraint support yes
no
Has acid support yes
no
Has metadata yes
yes
Has encryption support maybe
yes
Data processing framework support
Apache Spark,
Apache Flink,
Apache Beam,
Apache Drill,
Apache Flink,
Apache Spark,
Analytics query support
Apache Hive,
Apache Impala,
AWS Athena,
BigQuery,
Clickhouse,
Presto,
Trino,
Apache Hive,
Apache Impala,
Apache Druid,
Apache Pinot,
AWS Athena,
Azure Synapse,
BigQuery,
Clickhouse,
Dremio,
DuckDB,
Firebolt,
Apache Hudi
Apache Parquet