The speed and low latency of dynamic access memory (DRAM) are why so many organizations are looking at developing new applications to take advantage of the in-memory computing concept. The problem is that rewriting and migrating to a new application is no easy task. What if you could create an in-memory computing environment without rewriting your applications?
To make a seamless in-memory environment requires a software solution that can make traditional DRAM look like a block storage device. The challenge, of course, with using DRAM is it is not persistent, and while there are plenty of other memory technologies like flash SSDs, that are persistent, none provide the performance and latencies that DRAM does. A solution that leverages DRAM as its primary storage area has to account for DRAMs higher cost, lower capacities, and lack of persistence.
Introducing Formulus Black
Formulus Black developed a software-based storage software solution called Forsa, which is designed to make DRAM act like a standard block device. It offsets DRAM’s higher cost by leveraging a data efficiency technology that provides more significant resource optimizations than deduplication or compression. Formulus Black patented a data encoding technology called Formulus Black Bit Markers by providing real-time pattern matching. Formulus Black claims efficiencies of 2-24X. The Bit Marker technology combined with the steady decrease in RAM prices, as well as the introduction of technologies like Intel Optane, enable Forsa to be a cost-effective option for any organization concerned about current performance and latency limitations of flash.
Formulus Black’s BLINK technology accounts for DRAM’s lack of persistence by allowing data to persist across power cycles. BLINK enables the periodic real-time capture of Bit Markers and saves those to a more persistent storage medium like flash SSDs. The combination of BLINK and reliable UPS technology mitigates much of the persistency risk. In 2.0 a “BLINK” could be taken of the entire system and persist multiple terabytes of data from DRAM to local SSDs in just minutes (BLINK is further improved in version 3.0, below)
Forsa 3.0
Formulus Black recently released version 3.0 of the software. A significant component of the update is that Formulus Black, no longer exclusively couples Forsa to Ubuntu Linux. The new release supports a variety of Linux based operating systems including CentOS. Red Hat Enterprise Linux (RHEL) support is coming soon. Since Forsa includes Linux KVM Hypervisor support with each supported operating system, organizations with virtual machine workloads can easily migrate them to the KVM hypervisor to reap the benefits of Forsa’s performance.
Also in version 3.0, Formulus Black added support for Intel Cascade Lake CPU and Intel Optane DC Persistent Memory. The new release also supports much larger 4 and 8 socket servers and can pool together the physical memory across all sockets as storage for applications. Forsa also provides DRAM based memory error checking and recovering with patented technologies.
3.0 also takes a step toward addressing capacity concerns. In the new release, the software can pool memory-based storage, across servers using NVMEoF/RDMA. The increase in capacity does add network latency, but for organizations needing capacities beyond the capabilities of a single server, it is a viable trade-off.
Before 3.0, “BLINKing” meant the entire storage pool had to pause while the BLINK occurred. In version 3.0, administrators can isolate a BLINK to a single virtual machine or LUN.
The use cases for Formulus Black’s Forsa include storage-intensive database workloads, high-performance computing, artificial intelligence, and machine learning. As an example of the impact of Forsa, Formulus Black cites a recent case study where the solution showed a 68.7x improvement in query performance for exploratory analytics versus the competitive system, a system with NVMe SSD storage media.
StorageSwiss Take
Organizations are finding a polarizing set of use cases in their data centers. A mid-range all-flash array can satisfy the performance demands of most of the organization’s applications. On the other end of the spectrum is a smaller set of applications than need all the performance that the infrastructure can provide. While smaller in number, these applications tend to be the ones most directly tied to revenue, innovation, and overall competitiveness. These applications need extreme performance.
Delivering extreme performance, so far, has meant building the most powerful flash array possible by combining flash with the most powerful processors and even proprietary field-programmable arrays (FPGA). Forsa changes the game, by leveraging the fastest and lowest latency storage media available, DRAM. Its software enables organizations to exploit DRAM while insulating them from DRAM’s cost and protecting them from its risks.
George Crump is the Chief Marketing Officer at VergeIO, the leader in Ultraconverged Infrastructure. Prior to VergeIO he was Chief Product Strategist at StorONE. Before assuming roles with innovative technology vendors, George spent almost 14 years as the founder and lead analyst at Storage Switzerland. In his spare time, he continues to write blogs on Storage Switzerland to educate IT professionals on all aspects of data center storage. He is the primary contributor to Storage Switzerland and is a heavily sought-after public speaker. With over 30 years of experience designing storage solutions for data centers across the US, he has seen the birth of such technologies as RAID, NAS, SAN, Virtualization, Cloud, and Enterprise Flash. Before founding Storage Switzerland, he was CTO at one of the nation's largest storage integrators, where he was in charge of technology testing, integration, and product selection.
StorageSwiss.com - The Home of Storage Switzerland 