Latest News

cality Announces 1 Terabit Per Second Performance for Scale-Out File System in Microsoft Azure Public Cloud

Scality, a leader in software solutions for distributed file and object storage, today announced the technical preview of its high-performance scale-out file system (SOFS) in Microsoft Azure. Scality SOFS running in Azure delivers linear scaling performance for read- and write-intensive workloads. Scality SOFS performance running in Azure was measured at 1 terabit per second—the equivalent of downloading 50 high-definition movies per second!

On top of this record speed, the service is up to 10 times less expensive than other file services capable of delivering similar performance. Why? Because Scality SOFS leverages Azure Blob storage, giving customers access to limitless storage at only $0.0184 per GB per month. And, with Azure Storage Reserved Capacity, customers save over 30% off standard pricing.

Most other cloud file services sit on virtual machines and can only use capacity allocated to the local virtual machines on which they run. This local storage model is expensive and limited in its ability to scale out.

“In late 2019, we decided to port our proven SOFS code base to Azure,” said Giorgio Regni, CTO and co-founder, Scality. “The combination of customer interest in hybrid cloud or pure cloud use cases with some of Azure’s differentiated features, such as a single API for storage tiers and Azure Data Lake Storage (ADLS), enabled us to quickly deliver an integrated solution on top of Azure Blob storage.”

Scality SOFS running in Azure provides a completely stateless, POSIX-compatible interface supporting SMB 3.0 as well as NFS v3 and v4.1. The solution is hosted in a customer’s Azure subscription and connects to the customer’s Azure Blob storage accounts. Any number of virtual machines can be spun up on-demand to linearly scale performance, and SOFS tiers data across Azure Blob to optimize performance and costs. Leveraging low-cost virtual machines and limitless object storage allows SOFS running in Azure to offer blazing fast performance at a low cost that can scale to hundreds of petabytes of data. And because the solution supports ADLS Gen2, data remains in native Azure format and fully accessible by any Azure service for use cases including data analytics, machine learning and more.

Scality SOFS has been available as part of Scality’s industry-leading on-premises storage solution, RING, since 2013. SOFS is a feature-rich, POSIX-compatible file system with enterprise-level security and both synchronous and asynchronous replication to provide a complete solution for data centre recovery.

SOFS is used in daily production by hundreds of global customers supporting large-scale, mission-critical deployments for petabyte-scale use cases. Today, SOFS supports on-premises applications with high-aggregate performance requirements, such as massive-scale file systems (10 PB and 20 PB SMB shares), 1 PB per day ingest rates of logs, healthcare-critical deployments for medical imaging, recording and broadcasting of hundreds of simultaneous high-definition channels, and long-term asset preservation in national libraries. All of these applications can now be deployed in Azure cloud without modification.

Scality SOFS running in Azure builds on Scality’s multi-year collaboration with Azure, which has resulted in an Amazon S3 to Azure Blob translation solution available in Azure Marketplace, Zenko Connect, as well as support for the Azure Blob API as a connector to Scality’s on-premises storage solution, Scality RING.

“We are glad to continue our ongoing collaboration with Scality,” said Tad Brockway, Corporate Vice President, Networking and Storage, Microsoft Corp. “Their ability to rapidly port Scality SOFS to the Azure platform speaks to the advantages Azure is able to offer our many ISV partners.”

Scality SOFS running in Azure is currently available as a technical preview. To learn more and to request additional information, visit Scality’s Azure Solution page at https://www.scality.com/partners/azure/.

 

About Post Author