Install hbase with Homebrew, Nix

Project links

• Homepage
• Repository
• Upstream docs
• Package manager page

Commands and aliases

• hbase
• start-hbase.sh
• stop-hbase.sh

Project history

The official reference guide's history appendix begins with Google's 2006 Bigtable paper, then places the start of HBase development at the end of 2006. The project was designed to bring Bigtable-like capabilities to the Hadoop ecosystem rather than to be a standalone relational database.

HBase became a Hadoop sub-project in 2008 and an Apache top-level project in 2010. That transition matters historically because it moved HBase from an add-on in the Hadoop orbit to an Apache project with its own project management committee, release process, community, and operational identity.

The HBase website describes the system as a distributed, scalable big-data store for random, real-time read/write access. It emphasizes billions of rows, millions of columns, consistent operations, automatic failover and sharding on Hadoop/HDFS, and APIs including Java, REST, Thrift, filters, and Bloom filters.

The reference guide is unusually central to the project history: it covers standalone quick starts, distributed modes, configuration files, shell usage, schema design, MapReduce integration, security, architecture, backup/restore, replication, APIs, performance tuning, operations, and development. That breadth reflects HBase's evolution from a storage engine into a production database system with a large operator surface.

Adoption history

The official Powered by Apache HBase page lists user-submitted deployments across companies, institutions, and projects. It includes early production or large-scale use cases from Adobe, Facebook Messages, Flurry, HubSpot, OCLC WorldCat, OpenLogic, Trend Micro, Twitter, WorldLingo, Yahoo, and others.

The adoption examples show why HBase became important in the Hadoop era: teams needed low-latency reads and updates, wide sparse rows, time-series or event data, search and analytics backends, and MapReduce/Spark-adjacent processing without abandoning the commodity-cluster model.

HBase's adoption was also ecosystem-driven. Its value came not only from the database itself, but from living near HDFS, Hadoop MapReduce, ZooKeeper, Java clients, REST/Thrift gateways, and later higher-level systems such as Trafodion or Spark integrations.

How it is used

The quick start uses a standalone instance where the Master, RegionServer, and ZooKeeper daemon run in one JVM, then connects through the hbase shell to create, list, describe, put, and scan a table. Production usage moves the same model into distributed deployments with separate daemons and HDFS storage.

Administrators configure HBase through files such as conf/hbase-site.xml and conf/hbase-env.sh, start and stop services with scripts such as start-hbase.sh and stop-hbase.sh, and use the shell, Java API, REST, Thrift, MapReduce jobs, and operational tools to manage data and clusters.

Why package nerds care

HBase is a classic package-manager stress test because installing the package is only the beginning. The formula has to deliver shell commands and scripts, but the real system depends on Java compatibility, Hadoop/HDFS behavior, ZooKeeper coordination, configuration files, daemon management, and careful version matching.

For package nerds, HBase also represents the Hadoop-era pattern where a local package can be used for development or a single-node quick start, while production meaning lives in clusters, configuration, and service orchestration. That makes its package metadata deceptively small compared with its operational footprint.

Timeline

• 2006: Google publishes the Bigtable paper.
• 2006: HBase development starts near the end of the year.
• 2008: HBase becomes a Hadoop sub-project.
• 2010: HBase becomes an Apache top-level project.
• 2020s: The official site continues to present HBase as a top-level Apache project and Bigtable-like Hadoop database.

Related projects

• Google Bigtable is the direct design reference named by the HBase documentation.
• Apache Hadoop, HDFS, ZooKeeper, MapReduce, Spark, REST, Thrift, and Apache Trafodion are related technologies in the official documentation and adoption pages.
• HBase's source repository is managed by Apache through Git/GitBox.

Sources

• https://hbase.apache.org/
• https://hbase.apache.org/book.html
• https://hbase.apache.org/powered-by-hbase/
• https://hbase.apache.org/source-repository.html