The explosion of Big Data has driven the need for low cost, scalable, analytics solutions. However, the performance of some analytics tools such as NoSQL databases has struggled to deliver insights in the turnaround time demanded by today’s business environments. SanDisk’s new ZetaScale™ software enables application developers to scale their applications by taking full advantage of the high performance capabilities of flash technology.
“Flash technology can deliver astonishing performance gains for enterprise and cloud applications without forcing customers to make any other changes to their existing data center environments. However many of today’s analytics, database and in-memory compute applications haven’t yet been optimized for flash storage,” said John Scaramuzzo, senior vice president and general manager, Enterprise Storage Solutions at SanDisk. “ZetaScale represents a major step forward in our vision of the flash-transformed data center—empowering software developers to scale and enhance their applications to meet today’s big data and real-time analytics demands, while lowering TCO.”
Ushering In a New Era of Data Center Application Performance
ZetaScale software enables rapid development of analytics applications, such as NoSQL databases and in-memory computing environments that take full advantage of the performance and lower cost of flash. By combining ZetaScale software, SSDs, and NoSQL or in-memory computing applications, customers can analyze larger datasets at near-DRAM speeds for a significantly lower cost and at higher data densities per server node. This opens up new opportunities and accelerates the performance of big data and real-time analytics applications in enterprise data centers:
· NoSQL databases such as Cassandra can achieve four times higher performance when used with SSDs vs. when they are running on hard disk drives (HDDs). When NoSQL applications are combined with ZetaScale software on SSDs, they can reach nine times higher performance MongoDB applications using ZetaScale software can deliver 20 times the performance vs. running on HDD storage alone and more than three times greater performance vs. utilizing SSDs alone.
· For in-memory computing applications, like fraud detection, supply chain optimization and risk analysis, ZetaScale software bridges the gap between DRAM and flash storage. This allows the working dataset to expand beyond limited and expensive DRAM capacities into terabytes of lower cost flash storage, improving Transactions Per Second (TPS), per dollar, by up to four times.
In making this announcement, SanDisk and its partner, GigaSpaces Technologies, also unveiled GigaSpaces XAP MemoryXtend, a solution that leverages SanDisk ZetaScale technology to connect in-memory applications directly to the flash on a system, enabling customers to process larger datasets in near real-time.
“Our customers are constantly looking for ways to cut costs without compromising on the extreme performance of their in-memory compute applications because their success can often be directly tied to the bottom line,” said Zeev Bikowsky, CEO at GigaSpaces Technologies. “ZetaScale makes it possible for our XAP MemoryXtend solution to connect applications directly to the flash on the system, enabling our customers to get more value out of the data they analyze and better maximize their budget by being able to leverage more cost-effective flash.”
Designing Applications with Minimized Latency
ZetaScale software provides a key-value object store that increases application performance by enabling them to move beyond block access inefficiencies common in the storage access layer. The software also provides optimized threading to maximize concurrency and minimize latency, as well as support for atomic operations, transactions and snapshots. Additionally, ZetaScale is compatible with any brand of PCIe, SAS, SATA, DIMM or NVMe connected flash storage device, providing customers the ability to choose, avoiding hardware vendor lock-in.
A ZetaScale software development kit (SDK) containing the software library and APIs needed to optimize NoSQL and in-memory computing applications for flash storage, is available now on negotiated terms to qualified developers.