Overview of Cloud Bigtable

Cloud Bigtable is a sparsely populated table that can scale billions of rows and thousands of columns, which can also store terabytes or even petabytes of data. A value indexes each row; This value is called the row key. Cloud Bigtable is suitable for storing extensive single keyed data with minimal delay. It supports high read and write performance at low latency, and is an excellent data source for MapReduce functions.

Cloud Bigtable is ideal for applications that require very high performance and scalability for unstructured key/value data, where each value is typically no more than 10 MB. Cloud Bigtable also excels as a storage engine for module MapRoot operations, stream processing/analytics, and machine learning applications.

How Google Cloud makes it cheaper?

Cloud Bigtable is exposed to applications through a number of client libraries, including the extension for support for the Apache Hepbase library for Java. As a result, it integrates with the current Apache ecosystem of open source Big Data software.

Cloud Bigtable has long been a huge, petabyte-wide analysis and operational workload for Google Cloud as a fully managed NoSQL database. At 0.65 per hour and node, this was never the cheapest service to run, especially for Google Cloud production workloads that implemented at least three nodes per cluster. Today, however, it is changing, and now you can run the Bigtable product workload on the single node.

Google Cloud Bigtable product manager Sandy Ghai said: “We want Bigtable to be an excellent home for all of your key-value and wide-column use-cases, both large and small.” 

Suggested read – Google Cloud Introduces A Managed Memcached Service

With this, Google Cloud now enables the ability to use replications for more sweeping availability for these smaller clusters, as well as the ability to quickly convert a node development instance to a node production instance as needed. Additionally, the SLA of the service now includes all Bigtable instances, no matter what their size.

It is pleasing to see this push by Google Cloud to bring smaller workloads onto the Bigtable, mainly the company’s current focus on large enterprise customers and their specific needs. But a company that only needs a node today could be the most significant cluster of companies in the future, and Bigtable’s will at least always be a hindrance for small companies to enter – and once a company has its challenges with a particular database service, it’s unlikely to change anytime soon.