apache-drill mit Homebrew, apt, Nix installieren

Projektlinks

• Homepage

Befehle und Aliase

• auto-setup.sh
• drill-am.sh
• drill-conf
• drill-config.sh
• drill-embedded
• drill-localhost
• drill-on-yarn.sh
• drillbit.sh
• runbit
• sqlline
• submit_plan
• yarn-drillbit.sh

Projektgeschichte

Drill was created in the Apache ecosystem as an implementation inspired by Google's Dremel paper and entered the Apache Incubator in 2012. The project graduated as an Apache top-level project in November 2014, with ASF materials emphasizing both the Dremel lineage and the goal of letting analysts query large, changing datasets directly.

The 1.0 release followed in May 2015. By then Drill was positioned as a self-service SQL layer for Hadoop, HBase, NoSQL, and files, with ordinary SQL clients, JDBC/ODBC tooling, and the bundled sqlline shell acting as the day-to-day interface.

Adoptionsgeschichte

ASF's top-level project announcement described a community with contributors from MapR, Hortonworks, Pentaho, Cisco, and others. Its adoption pitch was less about replacing a warehouse and more about shortening the path from raw data to exploratory SQL, especially where schemas were late-binding or data lived outside a single database.

Drill's package-manager appeal comes from that same shape: install a JVM-based query layer, point storage plugins at local files, HDFS, S3, MongoDB, HBase, Hive, Kafka, and similar systems, then query them from a shell or BI client without running a separate ingestion pipeline first.

Wie es verwendet wird

The Homebrew package exposes Drill launch and helper scripts such as drill-embedded, drill-localhost, drillbit.sh, and sqlline. In practice, package users tend to start with embedded or localhost mode for local file and object-store exploration, then move configuration into drill-override.conf for distributed or secured deployments.

Drill configuration is split between the HOCON-style drill-override.conf and environment overrides such as drill-env.sh. Apache's docs state that Drill loads these from /etc/drill/conf when present, otherwise from $DRILL_HOME/conf, which matches the Unix paths useful to package maintainers.

Warum Paket-Nerds sich dafür interessieren

Drill is interesting to package nerds because it ships as a large Java data system but still offers a concrete CLI workflow: install it, run sqlline or drill-embedded, and query local JSON, Parquet, CSV, or remote systems through storage plugins. The package boundary matters because Java version, config directory, launch scripts, and optional native clients can all affect the first five minutes.

It also represents the Hadoop-era style of packaging: a tarball-style distribution with $DRILL_HOME, conf files, shell wrappers, and cluster-oriented defaults, wrapped by Homebrew into something developers can run locally.

Zeitleiste

• 2012: Drill enters the Apache Incubator.
• 2014: Drill graduates as an Apache top-level project.
• 2015: Apache Drill 1.0 is announced.
• 2025: Apache Drill 1.22.0 appears in the official release notes index.

Related projects

• Google Dremel is the research system that inspired Drill's interactive, nested-data SQL direction.
• Apache Calcite, Hadoop, Hive, HBase, Parquet, Kafka, MongoDB, and cloud object stores are common neighboring technologies in Drill deployments.

Quellen

• https://drill.apache.org/blog/2015/05/19/the-apache-software-foundation-announces-apache-drill-1.0/
• https://drill.apache.org/docs/start-up-options/
• https://incubator.apache.org/projects/drill.html
• https://news.apache.org/foundation/entry/the_apache_software_foundation_announces66