Workload automation management has never been more important. As organizations integrate applications in and beyond the enterprise, workload automation managers must master the complexity of simply and securely managing this complex set of disparate services. But it goes beyond that. Workloads are becoming more intertwined with business services, increasing the need for dynamic scheduling and self-service capabilities at the consumer level. These new requirements drive the need for underlying tools that dynamically manage these workloads. Clearly, these tools must embrace what already works today—while accommodating what’s needed tomorrow—without disrupting operations.
A Look Back
The history of job scheduling has mirrored the history of corporate computing. As time progressed, “batch” processes became increasingly interdependent. Processing one part of a workload or job is now dependent on the completion and results of a different, previously run job or workload.
Adding to the complexity is the growing number of information technology processes required to support the increasing volume of business services that need to be delivered. A simple credit approval process, for instance, requires executing and managing several IT processes and workloads. Organizations need a single point of control or management for all the business services at a macro level and the underlying IT processes and workload. We need to manage the work holistically to deliver the services that business relies on. The IT department must be able to dynamically deliver business services that meet Service Level Agreement (SLA) requirements and be able to adjust to meet unexpected and expanding volumes of service needs based on changing market and customer requirements.
Delivering Business Service Dynamics
As workloads increasingly become intertwined with the delivery of services, several things must happen:
• Organizations must fully understand the scope of their workloads (across the entire enterprise) to fully grasp what they’re delivering and what business services the workloads are enabling.
• Workload tools must accommodate changes in schedules in real-time, responding to shifts in demand, technology resource availability, and triggers from various sources.
• Workload products must be able to schedule and manage processes across mainframe, distributed, and cloud-based resources. In doing so, the workload product should offer flexible management, enabling the work to move across platforms to accommodate outages, errors, and unplanned needs.
• Workloads must be self-service-enabled so users can quickly and easily schedule work to address evolving business needs without contacting the IT organization.
• As workloads deliver more business services, they must be managed within easily understood SLAs; the IT department should be notified promptly (ideally automatically) if things aren’t trending well and SLAs are at risk.
Taking It to the Next Level
Workload automation capabilities must become more tightly intertwined with the broader set of organizational business processes. Workload automation solutions must be flexible; they can no longer dictate a business process through rigid, rule-based structures.
Consider a business such as a neighborhood bank branch. Customer-facing bank personnel manage everything from customer deposits and withdrawal services such as insurance, brokerage services, and investment advice. The customer-facing bank branch is backed by the same IT infrastructure tasked with providing services such as payroll and cash management to corporate customers. These activities, whether direct customer-facing or occurring in a business-to-business relationship, are intertwined; they all affect the same back-office resources and data. This creates a need for real-time coordination and management within the parameters of the technologies and delivery platforms. Customer service as well as legal and regulatory responsibilities must be addressed, too. Handling all that is a tall order. It can best be achieved with a highly available, reliable workload automation solution that provides a single system of record and control—one that can simply and predictably coordinate all the various tasks and work.
Enter the Cloud
IT organizations increasingly need the flexibility to dynamically increase computing capability to respond to evolving business needs. One of the most effective ways to address this is to leverage the cloud for “burst” capacity, where additional resources are only provisioned and paid for when used. This also helps organizations manage spikes in demand in a predictable fashion while keeping overall operating expenses in line.
To most effectively use cloud resources, your workload automation technology must be able to dynamically detect the need for and provision cloud-based resources. This should be based on a business policy that lets organizations effectively trade-off the costs of the additional capacity against the penalties of missing an SLA. Once these resources are brought into play, the workload automation tool must then be able to seamlessly use these cloud-based resources side-by-side with on-premises resources to manage SLAs, assure security, and deliver the same level of management capabilities as would occur if only using on-premises infrastructure.
The Quest for Continual Improvement
Several years ago, studies were commissioned to better understand how organizations were committing their IT budgets. The studies yielded an 80/20 rule—many organizations spent about 80 percent on the maintenance and delivery of their ongoing operations (“keeping the lights on”) and the remainder on the delivery of new services and other innovation. As organizations look to their IT departments for innovative new services, they will need to shift from the 80/20 rule. Even small reductions on the 80 side of the ledger yield massive increases on the 20 side. For example, if you can reduce operational expense by five points (approximately 6 percent), you can increase the investment in innovation by a massive 25 percent (increasing investment spend from 20 to 25 percent). Even small, incremental improvements in operational productivity yield a significant amount of potential investment into new service delivery.
One of the best ways to improve operational models is by implementing continuous improvement programs. W. Edwards Deming, one of the fathers of modern statistical management and statistical process control, described a four-step approach to improvement: “Plan, Do, Check, Act.” This approach necessitates a simple structure for ongoing monitoring, simple implementation, easy change procedures, and the ability to implement changes quickly and consistently. As we examine how we implement this structure in the context of workload management, it’s clear that an organization would need to understand and manage capacity planning, workload lifecycle management, simple workload provisioning and integration with underlying domain management tools, such as network, system, and storage. Finally, organizations must be able to link their monitoring and management tools to their lifecycle management tools, ensuring these tools all work together seamlessly.
The importance of workload management shouldn’t be understated. The increasingly complex IT ecosystem mandates that organizations be able to leverage their workload tools to deliver more sophisticated business services. By aligning workload capabilities more tightly with the organization’s business processes, looking to the cloud as an option to enable flexible capacity when needed, and seeking continual improvement, IT organizations can maximize their investment in workload technology and make it work for the good of the business.