Growing the Scale of Analytics

GUEST BLOG BY TOM PARKER

Do you know where your data is? Is it archived? Deleted? How do you begin to harness the hidden value in your data to help you make better decisions and then use that data a gain a competitive edge? How do you capture information from unstructured data, rather than just a traditional database?

Organizations have long struggled with how to address the increasing volume and variety of data to extract information and insight. Apache Hadoop has been one of the focal points of this movement, with attention on how to run large-scale analytics on clusters of cost- effective hardware.

As organizations look to efficiently store and analyze massive volumes of data, we have crossed the point where a Hadoop cluster can be economically attractive. An increasing number of analytics and visualization products work with data stored in Hadoop. Adoption of Hadoop continues to increase. With this heightened interest, there is an increasing need to simplify big-data deployments.

Lenovo has taken our experience with big-data deployments and delivered a thoroughly tested and integrated solution, including certification of Hadoop reference architectures. By using one of these reference architectures, you can quickly deploy a proven design to get your analytics solution up and running.

We recently extended our big-data solutions from our System x M4 to our System x M5 servers, bringing with the transition hardware features designed specifically for big data. For example, the System x3650 M5 supports 14 LFF 3.5-inch disk drives for data, plus two SFF 2.5-inch drives to support the operating system.  With 6 TB LFF HDDs, this brings raw data capacity to 84TB per 2U chassis.

The M5 system runs on the latest Intel Xeon E5-2600 v2 processors and includes flash storage options and energy-efficient features. This makes it an ideal building block to serve as a data node in a Hadoop cluster.

Our reference architectures, built on Cloudera Distribution for Hadoop, MapR or IBM BigInsights, can easily scale, with a choice of infrastructure components including servers and networking. Lenovo System x servers have industry-leading reliability — a valuable feature for the name nodes in a Hadoop cluster. Recently, we were notified by ITIC that Lenovo servers achieved the highest reliability of all x86 servers in the industry for the third year in a row. With our servers, we also offer XClarity software, which helps reduce management challenges in a large cluster by assisting with auto-discovery and inventory and providing rapid-provisioning capabilities.

With this combined experience, reliable servers and management capabilities, building a high-performance, large-scale analytics cluster with Lenovo is easier than ever. For more information on our reference architectures, check out our big-data solutions.