Cloud Bigtable is Google's NoSQL wide-column database service similar in use case to Hadoop's HBase. It serves as the database that powers many core Google services such as Search, Analytics, Maps, and Gmail. It is a compressed, high performance, and proprietary data storage system built on top of a few Google technologies such as the Google File System, Chubby Lock Service, and SSTable.

Google describes Bigtable as a sparsely populated table that can scale to billions of rows and thousands of columns. Bigtable was designed to support applications requiring massive scalability and was intended to be used with petabytes of data. The database was designed to be deployed on clustered systems and uses a simple data model, which is the wide column store.

Data is assembled in order by row key, which is a single value in each row, and indexing of the map is arranged according to row, column keys, and timestamps. It is also sensitive to hot spotting and hence the key structure design has to be done carefully. According to Google, Cloud Bigtable is ideal as a data source for MapReduce operations and a storage for very large amounts of single-keyed data with very low latency, as it supports high read and write throughput at low latency. BigTable is exposed to applications via a gRPC API and an HBase-compatible API in Java.

There are several key advantages of BigTable over a self-managed HBase installation:

• Scalability: You can scale your cluster up to handle more queries by increasing your machine count, as opposed to a self-managed HBase installation which has a design bottleneck that limits performance after a certain QPS is reached.
• Simple administration: Cloud Bigtable handles upgrades and restarts transparently, as well as automatically maintaining high data durability, unlike HBase which requires managing masters, regions, clusters, or nodes.
• Cluster resizing without downtime: The size of your Cloud Bigtable cluster can be increased for a few hours to handle a large load, then reduced again, all without any downtime. It typically takes just a few minutes under load for Cloud Bigtable to balance performance across all of the nodes in your cluster, after you change a cluster's size.