Dynamically Manage Workloads with IBM Solutions

How can enterprise IT optimally manage workloads for best business value?

Tivoli Beat - A weekly IBM service management perspective. This question—a very familiar one to almost any IT manager—has no simple answer. In part, this is a consequence of infrastructural complexity; when the number of different technologies, platforms and management tools responsible for workloads increases, the challenge of managing those workloads increases as well. And unpredictable changes in business circumstances, or infrastructural failures or errors, will all too often turn a lens on that problem—amplifying it even more.

The outcome, of course, will be problematic utilization of the infrastructure in the pursuit of rapidly-changing business goals. For example, when demand is low, hardware idle time may still be high, creating unwanted costs; meanwhile, when demand is high, service levels may not always scale in proportion, resulting in diminished service levels for users, clients or customers.

Ideally, organizations would be able to leverage the infrastructure more dynamically—reducing the costs and complexity associated with workload management, while increasing workload scalability and flexibility. Such an approach would deliver many compelling workloads benefits.

Workloads would become more predictable, for instance, enabling quantified, calendar-driven forecasting based on business strategies and goals. Managing workloads would become simpler, as well, because those workloads could be managed from a single point of control, rather than separately on a platform-by-platform basis. And ensuring that the terms of service level agreements (SLAs) are met would naturally become easier; through business-prioritized, intelligent automation, the infrastructure could fulfill SLAs more quickly, more cost-effectively, and yet require less human oversight in the process.

IBM Tivoli Workload Scheduler delivers best-in-class workload management

Organizations that have invested in an IBM System z mainframe—thanks to the System z's industry-leading performance, security, scalability and reliability—already have the ideal platform to serve as the central point of control for optimized workload management of just this type. With IBM Tivoli Workload Scheduler for z/OS, they have exactly the tool they need to leverage the System z for this purpose.

Developed directly in response to IBM Tivoli customer needs and requests, this powerful offering consolidates and enhances workload management even for exceptionally distributed infrastructures, allowing workloads to be managed not on the basis of technical details, but, more abstractly, on the basis of business policies designed to lead to an optimized outcome. And in the latest version, IBM Tivoli Workload Scheduler for z/OS 8.5, many new features translate into an even more intelligent, powerful, cost-effective and efficient outcome for organizations today.

Workload Service Assurance: Solving the SLA deadline challenge

Services are, of course, commonly subject to the terms of service level agreements that spell out the nature of services to be delivered, as well as associated variables such as performance, time of delivery, availability, security or costs. Organizations must strive to fulfill these terms as fully as possible; failure to do so can result in many unwanted business consequences, ranging from missed deadlines to fiscal penalties to brand damage to diminished revenues and market share.

For this reason, IBM Tivoli Workload Scheduler for z/OS 8.5 now includes a key feature designed to help SLA-conscious organizations: Workload Service Assurance. This feature helps to ensure that SLA terms are fulfilled by proactively forecasting the possibility that SLA deadlines will be missed, and then taking action to mitigate or even eliminate that threat—with minimal human intervention and oversight.

More specifically, when a workload is delayed and an SLA deadline may not be met, Workload Service Assurance automatically and dynamically determines the critical path that workload takes through the infrastructure. Armed with this information, it can then (via integration with z/OS Workload Manager Service Classes), give higher processing priority to the workload, substantially improving the odds the SLA deadline will be met after all. This applies to both fixed-time and fixed-duration SLAs.

Furthermore, a monitoring interface with color-coded views of critical jobs and their risk levels give IT teams the information they need to assess workload health with respect to SLA terms—in real time—and take manual action if that is deemed essential for best results.

End-to-end management via a single point of control

Another common issue involved in optimally managing workloads is infrastructural complexity. Because workloads are typically delivered using disparate technologies and platforms, obtaining holistic, end-to-end perspective and control on how they are performing can be very difficult to achieve.

IBM Tivoli Workload Scheduler for z/OS 8.5 includes IBM's answer to that issue: the Tivoli Dynamic Workload Console. This console is an elegant, unified point of control for enterprise-class workloads that includes monitoring, business prioritization, modeling and reporting and is bundled free of charge with the solution.

Thanks to the Web-based interface, IT teams can access this feature from any browser on the IP network, reducing deployment time and total cost of ownership while increasing management convenience and business responsiveness to changing conditions. Visual dashboards reflect workload health levels and enable teams to create and execute workload-related tasks via an intuitive interface. Modeling functions help show how jobs, jobstreams, workstations and other resources are contributing to overall workload delivery.

And all of this information can be aggregated, analyzed and incorporated into reports or other forward-looking initiatives. For instance, forecasts can be created to anticipate future developments and proactively make current workloads even more efficient and cost-efficient. And organizations interested in anticipating how well future workloads will probably perform, once put in production—via preliminary trials, tests and simulations—can now accomplish that goal as well.

Intelligent automation maximizes business agility while minimizing costs

Two more compelling new features are the enhanced event management and conditional dependency now included with IBM Tivoli Workload Scheduler for z/OS 8.5. These deliver exceptionally fast and consistent responses to emerging problems or developments.

Suppose, for example, that workloads should be triggered based on changes in files (whether z/OS, HFS, or distributed) that reflect dynamic events. IBM Tivoli Workload Scheduler for z/OS 8.5 now considerably simplifies the workload triggering process. Via an included file-watching utility, which runs continually in the background, file changes are detected; based on policies, the solution then executes predefined workloads until specific conditions have been met.

Conditional dependency features also deliver an intelligent, swift response to changing conditions through automation. In the event technical problems occur, generating error codes, the solution can now implement workloads conditioned on those codes (or combinations of them), helping to spur self-recovery of services and bring mission-critical workloads back into alignment with business goals. This response is governed by logical rules—essentially forming a tree of both problems and their optimized responses that helps to ensure that workloads meet business expectations.

Thus, IBM Tivoli Workload Scheduler for z/OS 8.5 improves workload performance, and reduces the many costs and complexities commonly associated with workload delivery. More generally, however, it enhances overall business resiliency even in the face of unpredictable demand levels—a key consideration in a problematic economy such as today's.