NoSQL databases store data differently than relational databases and come in several different “types” based on their data model. Because of the massive demand for data at almost all companies, NoSQL databases have seen robust growth in the past several years, and some estimates predict that the global NoSQL market will grow as much as 30% through 2025.
But the unexposed costs of NoSQL database ownership could create an unsustainable burden for many organizations. To help IT decision-makers understand the problem, GigaOm Analysts William McKnight and Jake Dolezal have published a new report, Cassandra Total Cost of Ownership Study, which examines the full costs and true value of the popular Cassandra NoSQL database self-managed on Google Cloud and the fully-managed DataStax Astra serverless Cassandra offering.
The final three-year Total Cost of Ownership (TCO) finds the cost of serverless Cassandra to be $740,087, compared to a cost of $3,174,938 for a self-managed OSS Cassandra deployment. It’s an astounding difference, according to the researchers.
McKnight says the serverless option “provisions what is necessary as it is necessary, removing the complex and costly overbuilding that occurs otherwise.”
He admits he was surprised at the extent of the cost savings possible with a serverless Cassandra deployment, which was found to be 76% less expensive than self-managed, with 95% less staffing costs, 3x less complexity, and 44% lower infrastructure costs than self-managed Cassandra.
“I have done less rigorous studies in the past for client situations comparing open source and commercial options and have come to similar conclusions, though the gap was not as pronounced as it was in this one,” McKnight says. “So I was surprised at the 44% lower operational costs for Astra versus OSS Cassandra.”
The analysts note that while cost is important, it is only one factor for a Cassandra platform selection. There are numerous other things to consider, including performance, administration, features and functionality, workload management, user interface, scalability, vendor, reliability, and numerous other criteria. But the research proves out that serverless Cassandra can significantly reduce deployment time and costs over self-managed Cassandra workloads. The report contains all of the information to replicate this test.