How Is Data Virtualization Different from Storage Virtualization
Posted in news
By Josh Miner, Primary Data Senior Director of Product Marketing
Storage virtualization technology has been around for about 20 years, but its capabilities have been limited to a single storage container, class, or vendor. The well-known phenomenon of data being trapped within a storage device is commonly referred to as a “storage silo,” and this problem is not solved with traditional storage virtualization. Enterprises today require a wide range of storage to meet different needs, from in-server flash and flash arrays, to SAN and NAS systems, to object and cloud storage. Siloed storage virtualization cannot cross these storage types, making it a less than compelling solution for modern datacenters.
Unveiled today, the DataSphere platform finally allows data to be moved easily between different types of storage. DataSphere is a storage agnostic solution that virtualizes data to solve a number of problems, including costly overprovisioning, imbalances between compute and storage, and the resource sprawl that results in an increasingly complex datacenter filled with numerous different storage devices.
Modern Technologies Increase Datacenter Complexity
While storage siloes have been a thorn in datacenters' sides for many years, two recent technology trends are creating new siloes that introduce more datacenter complexity and the urgent need for a solution. One of these trends is the explosion of computing endpoints, which is generating a huge increase in the amount of data enterprises have to process quickly. As a result, flash memory has emerged as a new high performance storage tier. Another trend is the adoption of cloud computing, which offers access to inexpensive and flexible resources, but at the cost of control. Let's discuss each of these in turn.
Dynamically Dealing with the Data Deluge
The current explosion of computing endpoints has created the need for in-server flash. From mobile devices to mobile apps to the Internet of Things (IoT), organizations are being inundated with data. While there is great value in processing this data quickly, doing so requires both high performance computing and fast storage. As data volumes increase, in-server flash on analytics servers will become a must. However, these flash-based app servers are a new storage silo that IT needs to overprovision and manage.
While flash is needed for real-time data processing, it is currently too costly to fill with the massive volumes of mostly transient data that is typically hot initially, but stale within hours or days. Even though this stale data is using up an expensive resource, it’s too difficult and disruptive to business for IT to move it off servers and onto more cost-effective disk or cloud storage. DataSphere can fix the problem of poor capacity utilization by making use of disk and cloud resources, while improving performance by using Tier 0 cache for application data, such as temp, log, and swap files.
DataSphere enables all resources to be shared within a global dataspace, including in-server flash. It is also data-aware, monitoring resource usage at the computing endpoint. This creates many exciting possibilities, both in storage and even in application architectures. For example, in analytics, instead of copying data to a separate platform (and storage silo), analytics might be performed in-place depending on resource availability at that moment, or alternately copied to other server flash to complete the entire copy and ETL process very quickly. After the job completes, clean up of transient data could happen automatically.
Siloes Curb Cloud Adoption
The advent of public and hybrid clouds is another trend introducing new silos into the enterprise. The cost savings and flexibility clouds offer enterprises are compelling. In fact, Gartner predicts that almost half of large enterprises will be engaged in a combined, public/private cloud operation by 2017. Still, the inability to easily move data between clouds and datacenters is creating new problems and inhibiting cloud adoption. Let’s take a closer look at several different use cases.
Simplifying Data Management with a Holistic Solution
In the last few years, cloud storage has evolved to a point where protection and cost levels are attractive enough for enterprises to consider using the cloud for archival. However, IT still has not been able to incorporate cloud storage as an integrated tier in the datacenter. Instead, the cloud has created another silo of storage that sits outside the datacenter.
The DataSphere platform virtualizes data across a global dataspace so that management spans both on-premise storage tiers, as well as public and private clouds. This enables enterprises to create holistic policies that address active data, archiving, and disaster recovery, whether data is hosted in onsite storage or off-site cloud storage. In addition, intelligent DataSphere data management capabilities offer snapshot-based data protection that includes seamless cloud archival of snapshots and stale data. This enables enterprises to maintain data protection with a single solution that reduces total cost of ownership by eliminating the need to purchase and manage separate archive, data migration and DR solutions.
Empowering Enterprises to Secure Data
According to the 2015 Cloud Research Partners' Cloud Security Spotlight Report, security is still the biggest barrier to cloud adoption, with nine out of ten organizations reporting concerns about cloud security. DataSphere resolves security issues two ways. First, it allows organizations to encrypt data before sending it to the cloud, and then spread the data by striping it across multiple cloud providers. This protects data by making the individual fragments meaningless in the event that an entity breaks the encryption for one of the cloud providers.
Enabling Cloud Cost Savings for Highly Regulated Industries
Another area where DataSphere can facilitate cloud adoption is in highly regulated industries, such as Finance, Healthcare, and Government. Compliance initiatives in these sectors often require that organizations provide fast access even to very old data quickly in the event of an audit. Since it can take days to retrieve archived data from the cloud, this means organizations lose out on much of the cost savings. For certain kinds of data—say the old high-resolution scans stored in a hospital's datacenter—this loss can be truly significant. DataSphere can quickly move data from off-premise storage such as S3 or remote Swift stores to on-premise resources, allowing companies to turn the cloud into an “active archive” to greatly reduce storage costs.
Stop Managing Storage and Start Managing Data
DataSphere is about much more than smart storage management; it's about mapping data intelligently to the needs of applications as those needs evolve. This dynamic data mobility allows IT professionals to stop spending time overprovisioning storage capacity and performance based on highly educated guesses regarding the long-term needs of applications. Instead, IT can finally get ahead of the prediction game by creating policies that specify Service Level Objectives (SLOs) that directly align with application’s needs, leveraging software to automatically place data on the best resource to meet these objectives, and easily scaling resources without the need to manually migrate data. As enterprises grapple with pressure to increase performance on one end of the spectrum, while reining in costs of data growth on the other, DataSphere will to help IT teams find balance across these demands to help their companies lead in innovation.