DataSphere Data Virtualization Automates Intelligent Data Management

DataSphere Data Virtualization Automates Intelligent Data Management

Posted in tech

DataSphere automates data management with machine learning to enable enterprises to seamlessly deliver higher service levels at lower cost. We do this by virtualizing data so that enterprises’ diverse storage resources can be put to work in the service of applications, just as server virtualization puts compute resources to work for apps.

With VMworld 2017 in full swing, it’s useful to take a closer look at how data virtualization works and the way it finally brings the benefits of virtualization to storage. If you are attending VMworld, stop by and see a demo in-person at booth 831.

How Data Virtualization Works

Data virtualization uses a metadata engine to separate the metadata (the data about the data) path from the data path. This enables data management to transcend storage silos so enterprises can manage data across heterogeneous storage resources including flash, shared storage, and the cloud. The following diagram illustrates how this works:


Figure 1 – DataSphere separates the metadata (control) path from the data path

Once data can be managed outside storage silos, software can move it, on demand, to meet evolving business demands. Cold data that suddenly gets hot can be moved to flash resources, cooling data to lower cost storage, and data that has become cold can be archived to the cloud.

The DataSphere metadata engine uses IT-defined objectives for performance, protection, and price to arbitrate storage supply with data demand. When data falls out of alignment with objectives, it moves to data to the lowest cost storage available that meets data requirements. This movement is transparent to applications, using live data mobility capabilities native to NFS v4.2 that maintain data integrity even when applications access data in-flight.

DataSphere makes movement decisions intelligently and with application storage awareness. It gathers metadata intelligence in real-time to understand how applications experience storage (for example latency, IOPS and bandwidth). It also collects telemetry on the data that applications access, such as which files are open, closed, with modified dates and times, as well as any other metadata. In addition, it uses machine learning to adapt to events, such as end of quarter reporting. DataSphere can “learn” if data used by internal applications becomes active during end of quarter reporting (for example) and proactively move that data to higher performance resources, and then back to lower cost storage once the quarter closes.

Data Virtualization Benefits

Just as server virtualization improved the service and levels of apps and efficiency of compute resources, while centralizing the management of applications and application environments, data virtualization streamlines data management to deliver a number of benefits, including:

  • Reduced costs. DataSphere automatically moves cold data off expensive storage resources, including moving data into the cloud. This dramatically extends the life of enterprises existing storage resources. It also gives IT a comprehensive view of individual and global resources, and the ability to seamlessly add new resources on demand—significantly reducing the need to overprovision.
  • Simplified management. DataSphere transforms the manual work of migrations, upgrades, and reactive fire drills into an automated, continual, and proactive optimization process. This frees hours of valuable storage admin time, eliminates the risk of human error, and creates a more powerful role for storage admins: the data admin.
  • Improved performance. DataSphere proactively identifies and automatically remediates hot spots. In addition, its high-performance architecture delivers the performance and scalability to support the management of billions of files. This enables enterprises to nimbly meet the challenges of rapid data growth.

It’s important to note that DataSphere benefits span both physical and virtualized environments. DataSphere enhances VMware Storage Policy Based Management, integrating seamlessly with vCenter to automate the movement and placement of individual VMs and VMDKs across heterogeneous storage resources. This means that VMDKs that unexpectedly get hot can be fixed automatically, without requiring the intervention of an admin to Storage vMotion a VM. This automation also allows cold VMDKs to be archived to object/cloud storage to free resources for active VMs.

DataSphere enables storage resources to be put to work in the service of apps in the same way that server virtualization freed the resources of idle CPUs. The benefits to enterprises are quite familiar to enterprise vExperts: lower costs, easier management, and higher performance. To learn more, visit our product page or connect with us a deepdive@primarydata.com to schedule a meeting or demo.



Contact Form

Channel Partner