Transforming Data Center via Process Automation
By CXOtoday Staff
Aug 21, 2007

CIOs and data center managers have begun to recognize the need to operate IT as a business, like maximizing process efficiency, minimizing risk, and mandating accountability. As a result, organizations are investing substantial resources to define repeatable processes, integrate point tools, enforce policies, and automate IT operations.

Most data centers today, however, accept reactionary processes, slow delivery times, and high management overhead as inevitable. Often data centers operations rely on scripts to interface their processes to various technologies, and use homegrown "request portals" for tracking internal requests. These are labor-intensive, fragmented tools, and don't allow for efficient automation or effective collaboration. Most importantly, due to their one-off nature they're often prime culprits in configuration drift - a leading cause of production downtime.

With process automation, however, organizations can drive more flexibility and collaboration in their processes. By standardizing on policies and processes, organizations can avoid user errors, which lead to costly downtime. Moreover, measuring process metrics allows organizations to optimize time to resolution and reduce overall operational costs.

Process automation enables data center managers to master the complexity of IT infrastructure administration and transform their data centers through immediate and continuous process improvement and automation.

The Storage Supply Chain

For years, IT organizations have invested in enterprise software to automate business operations - such as production, distribution, and customer care. Results have been impressive, marked by significant improvements in the quality and responsiveness of business execution. Surprisingly, the management of the IT infrastructure that serves these systems hasn't been subject to the same level of scrutiny. After all, cobbling various applications together via scripts not only takes time that IT typically can't afford to devote, but it also yields disjointed and inconsistent results that don't easily migrate across multiple platforms, accommodate workflow changes, or integrate a variety of organizations and personnel.

Though some early work has been done in IT infrastructure management, these workflows must encapsulate more than simply capacity provisioning. It's about managing service levels and showing accountability across the storage supply chain. These processes and policies deal with everything from replica management to backup and disaster recovery - all key components of the storage supply chain. As importantly, these processes aren't confined to the storage administration function - they cross operational and organizational boundaries.

Driven by evolving business needs, hardware and software changes are often dictated -usually by people or events outside the storage organization. Operational practices, in turn, must change to accommodate these dynamics. Moreover, diversity and change characterize today's storage infrastructures. Process automation offers organizations a way to address these challenges and efficiently manage the distributed process execution required by today's networked storage environments while also matching organizational realities.

Automating Storage Processes

Process automation tools provide organizations with a service-oriented technology for workflow enabling, automating, and integrating their storage operations. Single tasks or complex processes can be encapsulated for use and reuse by others across the supply chain. Users in different organizations can manage their respective portion of an end-to-end process through role-based delegation of policy definition, process definition, and execution control. All the while, on-the-fly policy decisions can be delegated to authorize users across the process, regardless of organization or geographic boundaries. Automated processes can also be organized and published to reflect the way individuals in the supply chain views their job.

Best of all, it's fast and flexible. From single tasks to complex workflows, solutions architected for process automation enable rapid and iterative deployment as well as incremental change to help ensure continuous process improvement. Besides, a number of these solutions provide visual tools, so that administrators can easily model and implement user and system tasks, policies, and workflow. These solutions offer a library of discrete tasks and pre-defined customizable automated processes that model industry best practices. With these tools, organizations can manage in-flight inputs, approvals, user tasks, and point tool automation in a reliable, repeatable, and efficient way.

Process automation can ease security concerns associated with managing IT processes across distributed platforms. These processes should be managed with the same mission-critical approach as the line-of-business applications they support. Communication between the automation engine and its managed elements must be encrypted to protect data. Administrators must be able to manage state of processes to enable restart, recovery, and rollback. In addition, a system-enforced and user-extensible audit history must be provided, which captures details of user tasks, processes, and automation parameters.

Where to Begin

It's not surprising that the key practices for successful process automation implementation mirror those found in successful Business Process Management (BPM) deployments. Whether deployed alone or in parallel with monitoring tools and other traditional approaches to managing processes across the storage chain, the most effective process automation initiatives rely on 3 key fundamentals such as - thinking big, starting small, and scaling fast. Organizations select a portion of a single, well-understood process, such as on-demand reporting of system status and then begin the cycle of automating, learning, iterating, and evolving.

Even if an entire process is not automated initially, it's important to consider the service path all the way to the application. This will not only provide a roadmap for future automation, but will also help in establishing priorities, measurement criteria, and definitions for key process management elements. Once the process and management model is mapped, highly repetitive operations such as data migration, storage provisioning, and database cloning can be automated before tackling dynamic processes. Over time, many of these operations can be extended to further streamline storage management.

The final step in process automation implementation is to leverage the knowledge gained from automation to optimize and change practices over time. In other words, if a current practice isn't appropriate for automation, then automating it will do little more than repeat the same inappropriate practice but with greater efficiency. Consequently, organizations must evaluate and model process improvements that are made possible through automation. For example, many organizations may find that provisioning storage on demand rather than forecasting and automating it helps them avoid any unnecessary buildup of over-provisioned storage.

Clearly, for organizations that want IT to operate more like a business, process automation represents an effective solution for increasing efficiencies all across the supply chain. As a platform for standardizing processes and automating workflow across storage operations, process automation not only help organizations set and continuously improve measurable goals, but also quantifiably improve the process to achieve those goals and enforce compliance with the policies that drive resource optimization and appropriate service levels.

By Yogesh Agrawal, senior director (Product Management) of Symantec