AWS, Azure and Facebook have long understood the best method for gaining storage efficiencies is with a disaggregated scaleout infrastructure. Now, the storage industry is catching on, undergoing a major transformation from traditional storage options to modern software-defined solutions that better serve private cloud requirements.
Previously, the three ways to deploy storage included:
- Traditional Storage Area Network (SAN) products, providing centralized storage with high utilization and enterprise features, but delivering low performance and high latencies
- Rack Scale storage products, providing high-performance storage over RDMA fabrics, which is very difficult to deploy at scale across an entire data center
- Direct-attached SSD storage, providing high performance but resulting in stranded capacity with low utilization
As hyperscalers realized the benefits of disaggregated infrastructure, they invested in the technology because they could offer customers the following benefits:
- Unprecedented scalability
- Lower total cost of ownership (TCO)
- Higher application performance
- Network flexibility
Now, Lightbits Labs is bringing this hyperscale cloud infrastructure technology to the masses in private cloud deployments.
Lightbits has pioneered and standardized a technology to enable enterprises, hyperscalers and private cloud providers the ability to deploy hyperscale cloud infrastructure using standard NVMe/TCP, a low-latency and high-performance protocol with end-to-end NVMe that uses vanilla TCP/IP networks and standard servers.
IDC Research Vice President of Storage Eric Burgener believes TCP will enhance the use of NVMe over fabrics. “The TCP implementation will start to open up much broader use of NVMe over fabrics,” Burgener said. “When customers wanted extremely high storage performance and low latency, they typically bought a PCIe in a server and let the application talk only to that server. As workloads increase, you start to outgrow how much storage you can get from a single server.”
Lightbits’ software-defined disaggregation solution helps enterprise private clouds, Software as a Service (SaaS) and Infrastructure as a Service (IaaS) providers save unprecedented amounts of time and money while enabling higher application performance and public cloud grade hyperscalability. Lightbits was first to recognize the opportunity created by combining newly-available affordable Flash solutions with high-performance standard networks.
Its LightOS® software, LightField™ storage acceleration card and SuperSSD are the first NVMe/TCP solutions to provide end-to-end NVMe over standard network and provide a Global Flash Translation Layer (GFTL) that gets the most out of the SSDs. Lightbits’ GFTL significantly boosts performance and delivers rich data services while enhancing the endurance of commodity SSDs.
Application data is now on the network
Infrastructure and network equipment providers are now an integral part of the application data layer. It’s more important than ever to maintain simplicity and support high-performance TCP networks.
Cisco Live San Diego
Lightbits is delivering high-performance, low latency and resilient storage via NVMe/TCP to enable efficient scaling to enterprises hyperscalers and public cloud providers.
Lightbits will demonstrate its hyperconverged, disaggregated storage solutions at the Investments Village, #213-L at Cisco Live. Lightbits Labs CTO Muli Ben-Yehuda will give a talk titled, “Disaggregation without Drama,” at the Investments Village Theatre on Tuesday, June 11 at 12:40 pm and Thursday, June 13 at 11:45 am. During these sessions, Muli Ben-Yehuda will discuss how enterprises need to reimagine storage by separating storage from compute to effectively hyperscale.
Learn how the Lightbits’ software defined disaggregated storage solutions enable simple storage disaggregation at scale to maximize utilization, improve performance, and reduce cost. To learn more about Lightbits Labs’ cloud infrastructure and storage solutions, visit www.lightbitslabs.com.