This gives businesses the IT flexibility, and the business agility, to deal with changing business conditions.
IBM’s InterConnect Conference in Las Vegas focused on digital transformation, which leverages hybrid cloud services to deliver advanced data services to business end-users. In an age when the focus is focused on software and cloud, it’s important to note that a new generation of software-enabled hardware infrastructure (SDI) is increasingly taking center stage. It will take time to be fully adopted, but many organizations are already piloting SDI – or deploying it for new projects.
Supporting a Scale-Out Infrastructure
It’s clear that IBM sees SDI and SDS (software-defined storage) as key leverage points in digital transformation, as noted by IBM executives at InterConnect. Software-defined augments IBM’s hybrid cloud strategy, as set forth by Robert LeBlanc, IBM Senior VP of Cloud Computing in his keynote speech.
For IBM, the emphasis on SDI and SDS is historic. For decades, the traditional, classic IBM focused on scaling up its systems to increase capacity and performance, including System z mainframes, Power Systems and System x x86 servers.
Today, customers are shifting to a software-driven, scale-out world in which distributed data and distributed networks of servers are the norm – often linking systems and storage made by many vendors. To add capacity and capability, customers create domains within scale-out infrastructure – not by piling up compute blocks one within a single cabinet or frame.
Software-Defined Infrastructure for Hybrid Cloud
Case in point: one InterConnect panel, moderated by IBM Storage Vice President Eric Herzog, explored service-provider’s adoption of SDS for scalable storage resources in Cloud Computing.
Managing virtual storage resources allowed these SPs to rapidly grow capacity, on an as-needed basis. It also improved on-line backup of data on cloud-resident resources. Executives from NaviSite, Tieto Computing Cloud, SilverPop (recently acquired by IBM) and Cobalt Iron, said software-defined storage (SDS) allows them to adjust dynamically cloud storage as user demand and business conditions change.
Looking across the industry, SDS adoption will build up over time, starting with early adopters, as hybrid clouds link private clouds with one or more public clouds. IBM’s systems competitors, and its partners, are seeing the SDS trend, too, adding to software-based orchestration offerings. Examples include EMC, Red Hat, Hitachi, and Veritas with offerings in the SDS space. IBM is already well-positioned in SDS, adding to the IBM Spectrum software suite for orchestrating workloads across hybrid clouds.
Scale Up to Scale Out
Software is at the heart of this technology change. Today, Hadoop clusters, and scalable Linux clusters are often used to run analytics, database and enterprise workloads that once might have run on scale-up servers. IBM has invested in SDI and SDS, and acquired companies focused on these technologies (e.g. Platform Computing, CleverSafe object-store software), aiming to be one of the top providers in these new market spaces.
As for hardware, entry pricing for IBM systems hardware has improved (e.g. IBM System z LinuxONE and IBM Power systems for Linux) allows customers to start small – and to choose between scale-up and scale-out options for supporting mission-critical workloads. This opens the door for net-new customers, especially those in service providers (e.g. CSPs and MSPs) and emerging emerging economies to run Linux workloads on System z and Power systems.
Hardware Abstraction, Software Orchestration
Hardware abstraction is key to the SDI strategy, allowing workloads to “flow” over multiple platforms, as long as they are riding on top of Linux, open source, and IBM middleware (e.g. IBM WebSphere). This is also applicable to IBM’s IaaS (SoftLayer), PaaS (BlueMix) and SaaS-delivered cloud services, strengthening IBM’s case in cloud-delivery of applications and workloads.
Scalability, reliability, availability, quality-of-service (QoS) and security are all top-of-mind for large IT organizations developing hybrid-cloud end-to-end workloads. These customers have been adding “pizza boxes” for hyperscale scale-out computing, but many of them are relying on custom software for scale-out orchestration. Now, the industry, including IBM, is designing SDS software to keep SLAs and uptime as high as possible.
Business and IT Strategies
As Cloud Service Providers and Managed Service Providers become a bigger part of the overall IT customer “mix,” IBM can optimize its go-to-market programs to appeal to SPs. SDI, which leverages the strengths of IBM’s orchestration and storage-management software, is likely to be one of the pillars of customers’ move to IBM-enabled hybrid cloud deployments.
In the world of SPs, capacity needs often jump in quantum-like fashion, adding large amounts of capacity quickly, in a timing sequence that is difficult to predict or plan. In a digital services, scale-out world, resources must be added quickly, without re-architecting the data center. Software-defined infrastructure speaks to that need.