Datadog launches tool to help firms cut cloud storage costs
Datadog has released a storage management solution designed to help companies optimise their cloud object storage costs, addressing challenges as data volumes and storage demands increase with the rise of AI and data-intensive workloads.
Cloud storage pressure
As businesses expand their use of cloud services, they can accumulate petabytes of data made up of thousands of storage buckets and millions of objects. The rapid growth is often driven by AI workloads, including large training datasets and logs from model inferencing. Managing this exponential storage growth has become a significant concern for operations teams, with inefficiencies leading to higher costs and the risk of reduced service quality.
The new Datadog Storage Management tool offers automation for analysing costs, monitoring storage growth, detecting anomalies and enforcing storage policies. It is currently available for Amazon S3 with support for Google Cloud Storage and Azure Blob Storage expected in the future.
Cost visibility
The product promises to provide granular visibility into object storage usage by analysing storage at the bucket and prefix level. Organisations will be able to identify cost drivers such as infrequently accessed, temporary or duplicate data, and pinpoint which workloads or teams are associated with uneven storage growth or spending spikes. This is designed to support the enforcement of data lifecycle and tiering policies with a unified view of cost, usage and metadata.
Automated recommendations are included, with the system suggesting opportunities to re-tier, archive, or delete data as a way to reduce expenses. Proactive monitoring is used to detect anomalies in storage patterns and alert teams to unexpected changes, which supports more responsive management.
AI workloads impact
The impact of storage growth from AI development is a significant factor, with many businesses reporting that data storage and processing now represent a larger cost than direct AI model training or inference. Issues such as difficulty identifying cost drivers within large, shared buckets and a lack of context across metadata or access patterns have slowed efforts to manage storage more effectively.
"For companies building AI products, data storage and processing is consistently the third-highest contributor to cost-greater than expenses for AI model training and inferencing," said Yrieix Garnier, VP of Product, Datadog.
Time savings
Identifying storage inefficiencies such as cold data stored in expensive classes or the build-up of non-current object versions is usually a manual and slow process. Datadog aims to streamline this by using automation and targeted alerts, saving team resources and supporting faster responses. Unified context provided by the system is intended to make the enforcement of lifecycle, retention or tiering policies easier for companies managing large-scale, multi-cloud environments.
Storage management capabilities will complement Datadog's existing cloud cost management offerings by providing a more in-depth focus on object storage. The company expects this will support organisations in tracking trends, identifying optimisation opportunities and implementing cost-saving actions more efficiently.
"With Datadog Storage Management, teams are empowered to optimize cloud storage costs and prevent unexpected spend. By rightsizing these costs, companies can keep their focus on building better products and bringing them to market," said Garnier.
