This article can also be found in the Premium Editorial Download "Storage magazine: Everything you needed to know about object-based storage."
Download it now to read this article plus other related content.
Convergence in the cloud
Less apparent, perhaps, is the potential of cloud technologies to enable storage convergence. In some cases, this might mean converging off-site data vaulting and on-premises storage using products from vendors like Ctera Networks, Nasuni, Riverbed and TwinStrata. Other vendors, like StorSimple, are delivering primary storage linked to the cloud behind it to efficiently handle the whole lifecycle of a piece of data. We call this approach cloud-integrated enterprise storage, where a single storage system can manage a piece of data forever. Tiering, snapshots, backup, replication, archive and more can all tap into the cloud, and the cloud service appears as a seamless part of primary storage. This effectively converges data management and eliminates any unnecessary data movement. The potential financial benefit from reducing data movement, and the elimination of redundant systems, connectivity and software, can be substantial.
One + One = One
There’s also plenty of interest in the physical convergence of coupling storage and compute closer together to cut down on complexity. Major storage vendors like Dell, EMC (with Cisco) and Hewlett-Packard (HP) all offer physically converged packages. For EMC/Cisco and HP, it’s about seamlessly integrating servers and storage into a single infrastructure “block” that can scale easily. Dell’s approach is similar, but with the added twist of aiding convergence with all-in-one features
Newer entrants, like Nutanix and Pivot3, provide convergence in the form of building blocks that combine storage and compute. Scale Computing will soon introduce a similar approach with compute and networking running on top of n-way clustered storage nodes to create a seamless data center fabric. That type of physical convergence could signal a massive change as physical connectivity and subsystem performance limits disappear. It could also make it easier to strategically plan resource utilization and growth.
A couple of years ago, most major vendors were pushing highly converged and highly integrated stacks of IT gear, setting off talk of a next-generation, mainframe-like superstack along with the inevitable vendor lock-in. Since then, the “stack” din has died down. Today, visions of convergence are more likely to intersect with broader virtualization and cloud trends, with less concern about vendor lock-in. Whether you turn to new or established vendors, look for convergence. The evidence is convincing and there’s clear value for data storage shops.
This was first published in June 2012