Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:


Lightbits Labs Launches Industry’s First NVMe/TCP Clustered Storage Solution

LightOS is the first NVMe/TCP clustered storage solution that protects against data loss and avoids service interruptions at scale. 

Today Lightbits Labs announced it has delivered the industry’s first NVMe/TCP clustered storage solution with the company’s LightOS technology.

With more companies moving away from direct attached storage (DAS), and with storage requirements typically growing far faster than compute requirements, both public-cloud providers and private-cloud builders are looking for ways to separate storage and compute so each can scale separately. One of the limiting factors to scaling disaggregated storage, however, is the need for high availability across clusters of storage and compute. Lightbits LightOS delivers the availability, flexibility and efficiency of hyperscale cloud infrastructure to on-premise data centers.

LightOS is now the first NVMe/TCP storage solution that protects against data loss and avoids service interruptions at scale. In the presence of server, storage, or network failures, LightOS maximizes operational efficiency, ensuring applications continue working in the presence of failures, and failover is handled automatically, keeping data fully consistent and available.

The average cost of IT downtime is enormous. According to Gartner, the average loss is $5,600 per minute. This can add up to more than half a million dollars lost per hour at the higher end.

At cloud scale, everything fails, so we built LightOS to capture all the benefits of disaggregated storage while providing exceptional durability and availability to prevent service disruption,” said Kam Eshghi, Chief Strategy Officer at Lightbits Labs.

For cloud-native applications, access to available, performant and easily deployable infrastructure is key to business agility and application experience, at any scale – regardless of whether you deploy five nodes, five hundred nodes, or thousands of nodes.

Installed on commodity servers in large-scale data centers, LightOS is optimized for I/O intensive compute clusters, such as Cassandra, MySQL, MongoDB, and time series databases. With end-to-end NVMe, LightOS delivers high performance and consistently low latency. The result is a 10X increase in storage reliability and 50% decrease in total cost of ownership (TCO). As a standard-based, target-only solution that does not require installing any proprietary software on the client side, the ease of deployment at scale is unmatched. LightOS is built to run in your existing data center, requiring no changes to application servers or your network infrastructure, so that deploying it at scale is a breeze.

As LightOS provides highly available, flexible, and efficient disaggregated storage, application teams can focus their efforts on developing new services to achieve their business goals, while lowering spend and delivering high performance.

LightOS provides the following additional benefits:

  • Software-defined disaggregated storage for cloud data centers delivering similar performance as direct-attached NVMe SSDs and up to a 50% reduction in tail latency.
  • Independent scaling of storage and compute, driving better infrastructure utilization and efficiency and providing unparalleled flexibility.
  • Standards-based NVMe/TCP block storage access protocol for application servers to storage servers communication.
  • LightOS clusters leverage standard NVMe 1.4 multipathing and offers data replications on storage target side, which provides transparency to the client with quick failover.
  • Distributed and durable cluster management with fast failover handling.
  • No single point of failure in data and control paths.
  • Multiple LightOS clusters can exist in the same cloud data center and can be easily aggregated and managed as one large cloud-scale block storage solution.

Sign up for our insideHPC Newsletter

Resource Links: