Datera Software Platform Analysis

by IT Analyst Firm Evaluator Group

Evaluator Group

Software-defined storage (SDS) was originally developed as a more flexible and scalable alternative to traditional storage arrays and to facilitate the transition to an all-flash environment. It has been embraced by large enterprises with hyperscale environments spanning virtual machines, containers, and bare metal that need flexibility and data portability.

IT analyst firm Evaluator Group evaluated the Datera Data Services Platform, the leading software-defined storage solution deployed on standard hardware from every major server vendor. Its scale-out architecture combines internal storage from multiple server nodes into a common pool of capacity, accessed via block or object storage protocols.

As data sets and storage requirements of enterprises have grown, enterprise IT have turned to SDS solutions like Datera to provide a cost-effective, storage-only solution that can deliver enterprise-level performance and reliability. While the system enables the use of many different storage node profiles, Datera’s customers by and large use SATA, SAS and NVMe flash media to rollout all-flash storage environments to achieve the highest levels of performance.

The Datera Product Analysis covers:

  • Application-driven storage provisioning with storage policies, data services for bare metal, virtual machines and containers and for SQL and NoSQL workloads.
  • Scale-out storage system overview with node counts, fault domains, and cluster management capabilities
  • How the platform is designed for data resilience, continuous availability, performance, and adaptability via lock-less coherency
  • Heterogeneous environment supporting any hardware and storage media workloads

While other SDS technologies have a similar scale-out architecture, their processes for handling the metadata associated with data placement and data coherency are quite different from Datera. Most employ a distributed lock management scheme that can increase latency by creating bottlenecks or increase the inter-node traffic required to synchronize changes across the cluster.

– Evaluator Group