Big Value From Software-Defined Storage Requires New Skills
As software-defined storage continues to emerge in the enterprise, the enterprise itself will need to become more capable.
Data storage is changing, mainly because the size and scope of the data we collect and manage is changing, growing, and increasing in value and complexity.
Most who have worked with data become accustomed to storage appliances and arrays: fully integrated solutions that offer convenient ways to store and retrieve data in various forms. The skills required by storage admins to operate these appliances, storage area networks, network attached storage, and more have generally fallen into the category of knowing how to pick the right storage systems to serve workloads and operate graphical consoles to make adjustments. But software-defined storage is different.
Somewhere in the early ages of today’s microprocessor-based technologies, manufacturers demonstrated the ability to create a communications modem in an industry-standard architecture (ISA) x86 computer purely using software. The only hardware required was an adapter for the serial port. We called them “winmodems” -- one of the earliest examples of software taking on the role of hardware. They weren’t especially flexible, nor were they popular among the early Linux crowd because the software was proprietary and only ran on Windows. But it wasn’t necessarily a bad idea.
The prevailing approach of storage manufacturers has been to do the opposite: combine all of the necessary software and hardware into a single, fully integrated product. For years, building appliances seemed like a logical way to minimize cost and deliver convenient solutions.
Only now are emerging storage vendors showing that solutions comprised of open source software on standard, commodity hardware actually make more sense. Hardware investments plummet while the levels of flexibility, agility, and nimbleness increase by orders of magnitude. It’s the same basic idea as a winmodem, but the intent is different: Instead of limiting flexibility to achieve lower cost, software-defined storage is designed to deliver them both. But operators need the right set of skills.
Technically Savvy Users A Must
Early storage appliances were built with neophyte users in mind -- business people who had little technical acumen. It seemed reasonable that the more control was built into the hardware, the less sophisticated the users would need to be.
The return to software control, independent of the hardware, does bring with it a concomitant requirement for users to be more sophisticated in their approach to software and systems. Even IT industry professionals who have been trained in the use of industry standard platforms and products will find themselves happily challenged as they migrate to an open-source-based, software-defined storage environment.
As software-defined storage continues to emerge in the enterprise, the enterprise itself will need to become more capable. The storage operations team and the systems operations team will ultimately become one unit -- one team with shared skills. As such, we can consider software-defined storage to be a true application of the DevOps mentality to storage.
Since most open systems today are built on a foundation of the Linux platform, those managing a software-defined storage environment will need to attain Linux skills. Other skills, including programming, scripting using languages like Python, and automation-related skills, will also be required, as will specific skills in the use of individual vendors’ products. These skills, which are foundational for DevOps practitioners, will be new for most existing storage administrators.
With most enterprise storage budgets remaining flat, software-defined storage provides many answers to the question: “What am I going to do with all this data?” However, development of skills in existing staff will be critical to success.
Ross Turk is director of product marketing in the Storage and Big Data business unit at Red Hat, responsible for marketing the software-defined storage portfolio of products based on Ceph and Gluster. Turk comes to Red Hat from Inktank, acquired by Red Hat in May 2014, ... View Full Bio
We welcome your comments on this topic on our social media channels, or [contact us directly] with questions about the site.