Learn About Scality RING’s Exabyte Scale, Multidimensional Architecture with Scality

Scality’s Giorgio Regni presented at Cloud Field Day 23, focusing on the Scality RING’s exabyte-scale, multidimensional architecture. Scality’s origin story stems from addressing storage challenges for early cloud providers, such as Comcast. They found that existing solutions weren’t meeting the demands of petabyte-scale data and the need to compete with large providers. The company’s core concept is “scale,” and their system is designed to expand seamlessly across all crucial dimensions. This includes capacity, metadata, and throughput, allowing them to scale each of these components independently.

Regni emphasized the RING’s disaggregated design, highlighting its ability to overcome common storage bottlenecks. The architecture separates storage nodes, I/O daemons, and a connector layer, enabling independent scaling of each component. He shared impressive numbers, including 12 exabytes of data currently in production and 6 trillion objects stored, with customers having billions of objects and applications using the system. The presentation also contrasted Scality’s approach to that of competitors like Ceph and MinIO, highlighting differences in metadata handling, bucket limits, and the flexibility of the architecture’s scaling capabilities.

Finally, the presentation covered the multi-layered architecture that supports various protocols, including S3, a custom REST protocol, and file system connectors. The architecture is based on a peer-to-peer distributed system with no single point of failure, supporting high availability and replication across multiple sites and tiers. It can manage different tiers, such as Ring XP, the all-flash configuration, and long-term storage. Scality RING also offers multi-tenancy and supports usage tracking, allowing customers to build their billing systems, with the overall goal of the system being an infinitely scalable storage solution.