Apache Calcite vs Apache Drill

Apache Calcite is a free, open-source data management framework for applications that need database-management components without built-in storage. It provides an industry-standard SQL parser, validator, and JDBC driver. Queries are represented as relational algebra, then transformed with planning rules and a cost model for optimization. Calcite can mediate between applications and several storage locations or processing engines, and can push computation to a source database. Listed adapters include Arrow, Cassandra, CSV, Druid, Elasticsearch, JDBC, MongoDB, Redis, Spark, and Splunk. Its JDBC driver supports local connections and remote connections using JSON over HTTP or Protobuf over HTTP. Calcite is available as source and through Maven, for API or self-hosted use. Build prerequisites include Git and Java JDK 8u220 or later, with Java 11 preferred. It does not store data; each source must be registered through a schema or adapter.

Apache Drill is a schema-free SQL query engine for Hadoop, NoSQL databases, and cloud storage. It can query raw data where it resides without first loading it, defining schemas, or transforming it. Its JSON data model supports complex and nested structures, and a single query can join data across multiple datastores. Listed sources include HBase, MongoDB, HDFS, Amazon S3, Azure Blob Storage, Google Cloud Storage, Swift, NAS, and local files. JDBC and ODBC drivers connect Drill to tools including Tableau, Qlik, MicroStrategy, Spotfire, SAS, and Excel; developers can also use its REST API in custom applications. Drill can run embedded on a laptop or in distributed mode on a server cluster, and the homepage lists Mac, Windows, and Linux. Documented security options include Kerberos, username and password authentication, Digest, authorization, and impersonation. The software is free under Apache License 2.0. A compatibility limit: official support for Hadoop 2 and Java 8 was dropped with Drill 1.22.0.