The ability to extract and actualize intelligence is vital to competing in today’s digital economy. However, many enterprises are struggling to efficiently collect, manage, and analyze growing volumes of data. Storing data poses a major challenge for architectures based on legacy technology, inspiring manufacturers to quickly innovate ever larger disk arrays, object storage, and cloud solutions to ease data growth pains. Yet storage is only part of the problem; effectively using data is another.
Today’s enterprises are looking to high performance computing (HPC) solutions to simulate weather, model space shuttles, perform genomic analysis, and other data-intensive tasks. To streamline these processes, Hewlett Packard Enterprise (HPE) has created a hierarchical storage management platform for HPC environments that not only supports escalating data workloads but can also optimize data migration and analytics.
The HPE Data Management Framework (DMF) enhances data accessibility and management resources based on a tiered data model. Data is generally separated into three categories: hot data – actively used, warm data – should be available but isn’t needed immediately, and cold data – can be filed away. DMF lets enterprises move data to the appropriate storage tier, which helps to reduce costs, streamline data processes, and support future innovation. DMF offers users key advantages:
- Cost management – Tiered storage architectures not only allow data to be placed in the best location for accessibility requirements, but also decrease the cost of storage by automatically migrating data to the appropriate tier. DMF can maximize storage resources by preventing cold data from tying up resources needed for hot data. These capabilities significantly increase data availability and storage efficiency, while minimizing total cost of ownership (TCO).
- Automated workflows – With DMF, administrators can stage data for a specific job by creating rules to collect the files needed for that job, and migrating them to the right place at the right time for processing. Once the data has been processed, administrators can move the results to the appropriate tier. Task automation reduces manual errors, eliminates script management and maintenance, and works to streamline data workflows.
- Scalable storage architecture – DMF enables users to introduce storage technologies by automatically migrating, validating, and moving massive datasets to new storage infrastructure. Automated data movement can occur in the background, with no impact on data access or application performance.
DMF allows enterprises to allocate their data based on service level and speed requirements, write policies to control data movement, and migrate data invisibly to users and applications in the background. These capabilities ensure that data is highly available and accessible, according to an automated schedule or written policy. Now, enterprises can operate with confidence knowing their data is stored cost-effectively, in highly accessible primary storage or in low-cost secondary storage devices such as tape, disk, and cloud. Data movement can be flexibly based on file size, file type, user ID, etc.
Purposefully designed for HPC Linux® storage environments, DMF is empowering users to rapidly access, utilize, and share troves of critical data. This comprehensive management architecture is key to simplifying data administration, improving workplace productivity, enhancing data resources, and reducing storage costs.