• 1951 – UNIVAC pioneers use of magnetic tape for storage
• 1993 – Severe Tire Damage is the first band to live stream
• 2007 – Netflix launches their streaming media business and finally,
• 2017 – Cisco & MapR achieve a SPEC SFS world record for streaming workloads
The Data Age
That’s some pretty impressive company. Why is this particular benchmark result so significant?Because data is the lifeblood of business. According to IDC, we’ve entered the Data Age: The world’s data will double every two years for the next decade. How companies store, manage, access, and deliver hypercritical data will determine winners and losers.But such volume of data cannot be easily managed with conventional systems. As data-driven businesses gain prominence, demand is growing for integrated solutions encompassing high-density storage, powerful compute, and improved scalability: • Enterprises are collecting and storing more data for longer periods to support predictive analytics.
• Data is classified by “temperatures” (hot/warm/cold) for dynamic data tiering.
• Decision makers want to access and analyze data for real time insights.
• But traditional storage arrays are expensive and don’t easily support large scale-out landscapes.
A New Approach to Software Defined Storage
That’s why cisco developed the UCS S3260 Storage Server, a modular architecture designed to deliver efficient, industry-leading storage for data-intensive workloads. And that’s why we’re partnering with MapR to provide a complete end-to-end software defined storage solution.
We Blew It Away
Our S3260 Storage Server offers the industry’s best performance in a compact 4RU form factor. It supports either Scale Up (expandable to 28 drives/node), or Scale Out (easily adding multiple nodes to your integrated infrastructure), and is a proven Hadoop platform. MapR-XD provides a highly reliable distributed data fabric with enterprise-grade features such as Global Namespace and Multi-temperature store.
SPEC SFS world record demonstrates the performance of the S3260 with MapR-XD.
SPEC’s benchmark suite is the standardized method for evaluating performance using file server throughput and response time. We ran the Video Data Acquisition (VDA) streaming workload because it simulates applications that store data acquired from temporally volatile sources such as surveillance cameras.
Cisco and MapR have been setting records since we published the industry’s first Big Data benchmark in 2015. But this was the first SPEC SFS benchmark on a HDFS (Hadoop file system) compatible platform. Together, Cisco and MapR blew away all previously published SPEC SFS benchmarks: