AWS Launches Cassandra Service
AWS today unveiled a serverless Cassandra service that ostensibly will give customers the scalability benefits of the NoSQL database but without the infamous management challenges that accompany it.
Appropriately named the Managed Cassandra Service (MCS), the new offering gives AWS customers the power of Apache Cassandra, most notably the ability of Cassandra to run a single database across multiple data centers around the world while maintaining very low latency and a high degree of high availability.
Those aren’t traits that are associated with all databases, but they are among the design points for Cassandra, which was originally created by a pair of Facebook developers in 2007 to power the social media site’s search function. The Java-based database was based on Google’s BigTable database and was widely adopted by organizations that need to store petabytes of data, like Netflix, CERN, Reddit, and Uber.
MCS implements the Cassandra version 3.11 API for the Cassandra Query Language (CQL), which enables customers to host their existing Cassandra applications on the cloud giant’s servers. MCS runs as a serverless service, so there’s no need for customers to provision any EC2 instances or virtual machines upon them to run the database.
No provisioning of storage is necessary with MCS, as storage is fully managed by AWS for the MCS service. Customers’ data is automatically replicated three times across muilptoiple AWS Availability Zones for durability, the vendor says.
MCS will let developers build applications with “virtually unlimited throughput and storage that can serve thousands of requests per second without the need for capacity planning,” AWS’s Danilo Poccia writes in a blog post today.
The new service unleashes the power of Cassandra while hiding its complexity, Poccia writes.
“Managing large Cassandra clusters can be difficult and takes a lot of time,” he continues. “You need specialized expertise to set up, configure, and maintain the underlying infrastructure, and have a deep understanding of the entire application stack, including the Apache Cassandra open source software. You need to add or remove nodes manually, rebalancing partitions, and doing so while keeping your application available with the required performance. Talking with customers, we found out that they often keep their clusters scaled up for peak load because scaling down is complex. To keep your Cassandra cluster updated, you have to do it node by node. It’s hard to backup and restore a cluster if something goes wrong during an update, and you may end up skipping patches or running an outdated version.”
MCS was one of many new services unveiled today by AWS CEO Andy Jassy during a marathon 2-hour-and-45-minute keynote address at re:Invent, the company’s annual conference. Jassy noted that the company had just about every other database on its cloud, including relational, document, time-series, graph, ledger, key-value, and in-memory. With MCS, AWS is adding the wide-column store to its database family.
“You should have the right tool for the right job…” Jassy said.