Industry Outlook is a regular Data Center Journal Q&A series that presents expert views on market trends, technologies and other issues relevant to data centers and IT.
This week, Industry Outlook asks Josh Epstein, chief marketing officer at Kaminario, about the storage industry today and what role composable infrastructure might play.
Industry Outlook: What is the current state of the storage industry?
Josh Epstein: As the forces of digital transformation reshape enterprise IT and the data center, the data-storage industry is undergoing its own transformation to keep up with evolving market needs. The rise of SaaS businesses and on-demand applications is driving the need for cloud-scale data center infrastructure that is optimized for dynamic performance requirements to support new user experiences and that enables the advanced analytics workloads driven by AI, machine learning and big data. As a result, modern IT organizations are seeking storage solutions that they can optimize for specific tasks while delivering the flexibility of the public cloud.
The rise of cloud-scale applications and the “SaaSification” of the enterprise are causing rapid displacement of traditional on-premises data centers. With as-a-service infrastructure models experiencing tremendous growth, IT organizations supporting these environments must think differently about how they procure, manage and optimize storage resources. This transition is the start of a new era for the storage industry.
IO: Do public infrastructure-as-a-service (IaaS) offerings have any limitations?
JE: Although some businesses are moving certain workloads to IaaS platforms, many applications don’t lend themselves to general-purpose, multitenant public storage services such as AWS. In particular, large SaaS providers and consumer Internet-application businesses are seeking to provide a more cost-effective, controlled and secure set of infrastructure solutions than the hyperscale public cloud can provide. IaaS-based strategies offer enormous flexibility, but they lack the same control, security and economics at scale of private and bespoke cloud offerings. As a result, the industry is seeing a rise in enterprises and service providers that are building cloud-scale data centers to offer more-custom services with the same flexibility as IaaS. Composable storage infrastructure is an important tool for these forward-thinking IT organizations.
IO: What is composable infrastructure and why is it becoming a trend?
JE: Composable infrastructure is an innovative approach that enables private clouds to optimize for web-scale applications while delivering the agility and operational simplicity of the public cloud. By integrating a composable storage architecture with higher-level data-orchestration platforms, storage operations become extremely efficient, giving IT organizations performance and efficiency at scale while keeping storage-management resources simple.
Through composable infrastructure, data center managers will be able to dynamically create and reconfigure storage solutions at the software layer with no physical movement or hardware. A composable storage architecture provides the performance and efficiency of scale-up storage with the operational simplicity of scale-out storage.
IO: How are analytics-intensive applications transforming the data center?
JE: Cloud-era applications require an incredibly fast, intelligent and flexible data infrastructure to maximize their potential. As such, data centers must be equipped to enhance the performance of on-demand applications and boost IT productivity. By employing technologies such as application-level intelligence, machine learning and big data analytics, operators can ensure more productivity out of the storage environment and deliver higher performance for business-critical applications.
- Machine-learning techniques applied to customer data deliver actual insights on workload placement, performance tuning and fabric configuration that directly increase business productivity.
- Automation technologies in the data center actively manage the unpredictable behavior of today’s cloud-era applications, allowing businesses to spend less time managing their infrastructure and more time solving critical business challenges.
- Using big data analytics and predictive modeling highlights actionable insights, simplifies storage management and brings the agility of hyperscale environments to enterprise and cloud providers.
IO: What is the state of automation in the data center?
JE: Although it once seemed futuristic, the autonomous data center is gaining traction. As the enterprise moves rapidly towards cloud strategies, it’s critical that data centers are operationally efficient. As a result, technical operators must think differently about how they build and optimize their facilities. Many businesses are automating data center operations to simplify infrastructure management. By integrating automation, operators will no longer have to monitor and adjust operations manually. What’s more, businesses will gain more insight to optimize system performance, adjust workflows and extend their physical infrastructure.
With regard to making the autonomous data center a reality, composable infrastructure is set to transform the enterprise storage industry as more organizations strive to replicate the performance and agility of the cloud. Software-defined, composable storage technologies can automate infrastructure changes on the basis of dynamic feedback from intelligent monitoring systems. Composable infrastructure platforms can be integrated into a broader set data center orchestration platforms.
IO: What are the benefits of composable infrastructure over traditional array-based architectures?
JE: Traditional storage arrays have forced customers into siloed storage infrastructures that can’t deliver the cloud-like agility that software-driven businesses require. Composable infrastructure allows enterprises to mix and match as well as shift workloads on demand in addition to consolidating workloads to increase utilization.
The composable paradigm promises to deliver a new level of agility to the data center by dynamically creating servers that meet individual requirements from logical resources that aren’t limited by physical constraints. Through the implementation of a composable storage architecture, today’s operators and service providers mitigate costs by providing software-defined consumption-based pricing without sacrificing performance, scalability or agility.
IO: What are the current cost dynamics in the storage industry and what do you forecast for the near future?
JE: As always, we see consistent price erosion at all tiers of the storage industry. This trend is at both the storage-media layer and the integrated-appliance layer. With NAND flash prices predicted to decline rapidly over the next several months, we anticipate a period of accelerated price erosion in the all-flash array space as well. Software-defined composable storage strategies let infrastructure providers get the maximum benefit from the rapid commoditization of storage media in addition to compute and networking hardware.