Switching From Traditional to Software-Defined Storage

By CIOReview | Tuesday, June 13, 2017

The ability to abstract the control plane from the underlying hardware has given rise to software-defined technologies. Software-defined technology brings in a lot of advantages, whether it is with software-defined networking and its ability to change at run-time or the software-defined data center, which pools together resources from the physical hardware and distributes them to applications as needed. Software-defined storage (SDS) uses abstraction to separate the storage services from the underlying hardware. Without being tied down to any physical system, storage resources can be used more efficiently, and automated policy-based management can help to simplify administration tasks. SDS is often associated with virtualized storage, but they have a key difference; SDS separates the storage capabilities and services from the hardware while virtualized storage decouples capacity. For building a storage system outside of the constraints of legacy solutions around the data center, the technology is moving towards an intelligent software layer embedded in commodity capacity nodes. Different forms of software-defined storage technology are mentioned below:

Single Console for Controlling Multi-Vendor Storage: Resembling an external storage virtualization and a re-branding of those same features and benefits under a more up-to-date moniker. Along with the potential benefits, the drawbacks are also associated with it. Newer offerings have added orchestration, automation, and API controls available to applications, especially in a cloud or virtual server environment that can offer storage control to the application or management layer, making a noticeable difference in terms of its functionality. Atlantis USX, DataCore and some iterations of IBM SVC are moving away from storage virtualization and aiming towards software-defined technology, offering better results through external influence and control.

Storage as software-only: Realizing the advantage of software over hardware, storage vendors are offering only the software, and integrating it with commodity hardware of the customer’s choice, proving inexpensive for enterprises using existing capacity. However, the downside is that customers have a bigger responsibility when it comes to architecting, testing and troubleshooting solutions compared to buying a fully integrated hardware and software combination. Most of the object storage systems today, such as Scality, Cleversafe, and Ceph, along with open source storage such as Nexenta or OpenStack-based options, are available in software-only configurations and describe themselves as software-defined storage.

Converged systems: Vendors of converged systems, like Nutanix, SimpliVity, and Coho Data, tailor storage, and server combinations to be particularly effective at running specific workloads— like server virtualization and orchestration. In this case, the software defining the storage is a combination of their internal system software, and the virtualization software—that gets to control the storage.

Virtualization in Hyper-Converged Infrastructure

A hyper-converged system allows the integrated technologies to be managed as a single system through a common toolset. Under the converged infrastructure approach, a vendor provides a pre-configured bundle of hardware and software in a single chassis with the goal of minimizing compatibility issues and simplifying management. Hyper-convergence is developed from the concept of converged infrastructure. Unlike converged infrastructure, the technologies in a hyper-converged infrastructure are so integrated that they cannot be broken down into separate components although can be expanded through the addition of nodes to the base unit like virtualized workloads.

Virtual Volumes Implementing Virtualization

Virtual Volumes (VVOLs) are a vSphere 6 feature considered as one of the newest types of data storage virtualization technology. It is an integration and management framework for external storage that provides finer control on virtual machines. VVOLs allow storage capacity to be aggregated into what can be thought of as containers, to which VMs can be matched based on the amount of capacity and performance they need. By doing so, virtual volumes are delivering flexible and efficient storage by efficiently managing data in virtual environments.

SDS  is a profound new way to deploy storage services that promise to lower the cost of storage, improve performance, and increase flexibility—allowing functions such as creating a volume, establishing RAID protection, implementing thin provisioning, and storage tiering through a single interface.