Question 1:
Data center managers are transforming their traditional infrastructure to support upcoming growth and business demand. Data center modernization was discussed by almost 30 participants at an analyst-moderated workshop at the summit. It was revealed that traditional ways of designing, operating and managing data centers are becoming outdated, as they offer limited flexibility to make changes along the way. Power availability is increasingly becoming a concern in India, and is impacting the cost of operations.
A common thread in our client conversations about the intelligent data center is that users are struggling to conceptualize what terms like "software defined" mean in practice, and how to codify them into a standard approach when determining the value of their data center strategy, investment and operations. In essence, users are struggling with the basic question of what sort of approach should be taken when building a new data center or renovating an existing one.
Systems of record: Established, packaged applications or legacy homegrown systems support core transaction processing, and manage the organization's critical master data.
Systems of differentiation: Applications that enable unique company processes or industry-specific capabilities. They have a medium life cycle (one to three years).
Systems of innovation: New applications that are built on an ad hoc basis to address emerging business requirements or opportunities. These are typically short life cycle projects (up to 12 months).
Systems-of-record applications, preferably, should be kept in house, whereas, systems of innovation should be developed using off-premises cloud solutions. This approach will achieve dual objectives - improving data center agility, and reducing capital expenditure (capex) investments (including real estate, which is expensive in Tier 1 cities in India). However, as with all changes in data centers, this approach will have cascade effects, especially on the organization. Support and maintenance for systems of record and systems of differentiation are well-established practices in all IT shops. However, newer systems of innovation, especially those supporting newer mobile platforms and application types, demand a completely different support structure.
Future data center operations will focus on workflow, rather than on where the work is processed. IT will not control everything; enterprises will move more workload types off-premises, and, subsequently, have external service providers (ESPs) handle some of the applications and data for the enterprise. As intelligent data center evolves, application and workflows will be tightly integrated. As intelligence is added to networks, these workflows could move dynamically across domains, based on business rules or external events, creating a tightly knit hybrid data center, and potentially changing the way we look at vendor relationships.
Technology choices are a critical part of this exercise. It is advised that users maximize their investment in commodity-based technologies, and virtualization is one of the popular options to drive agility and efficiency. Customers can use other options, such as containers and bare metal, in their data centers. Proprietary investment (such as Unix-based servers) should be considered for specific business-critical applications.
This also means that IT teams need to move out of their silos and collaborate as one team. The data center manager has to collaborate with the security and risk, finance, and market teams. At times, the IT teams, especially the data center manager, will have to act as a portfolio manager, risk manager, marketing manager or finance manager. Regarding risk, the data center manager will be responsible for creating business continuity plans, especially in case of any natural disasters. Unfortunately, the current focus concentrates on managing day-to-day operations, which needs to change.
Indian enterprises have taken a cautious approach toward cloud, and often try to link it with cost savings only. Client interactions suggest that many recent pilots failed because customers expected cost benefits, which the provider could not demonstrate. Clients are advised to create a compelling business case by focusing less on cost savings, and more on business aspects, agility, scalability, improved customer service and innovation. On the cost front, shifting the cost structure from capex to operating expenditure also should be factored in while analyzing the impact of cloud.
The business case discussion should not be restricted to the IT team, and should include leaders like the CFO, chief security officer, chief marketing officer and line of business (LOB) heads. For example, a customer experience team running a customer feedback program can consider leveraging additional infrastructure capacity via public cloud. With this collaborative approach, there are better chances of creating internal consensus toward considering IaaS in a strategic option for procuring IT.
As much as these are essential components in management of IT operations, alone, they will not improve availability and performance. A detailed assessment is required to understand the capabilities or gap in any existing tools before deciding to procure a new tool.
Question 2:
Procurement of such tools should be directly linked to improvement in IT performance, as well as any visible and tangible benefits business can derive. During the procurement process, adequate consideration should be given to smaller tool providers, as they will be agile, cost-effective and provide better local support, and they are generally focused on resolving a unique issue. Another emerging option is a software-as-a-service-based offering, which brings an operating expense model to the operation.
Rather than implementing the asset management tool, Indian enterprise are advised to integrate it into the ITOM process. Asset management tools either take a back seat, or are often integrated in a central ERP tool, which could be SAP or Oracle. The reason is simple: Assets need to be identified and accounted for. However, these tools have no linkages with the real-time IT operations asset management tools, even if they are not connected to the core applications used by the finance team. Sometimes, asset inventory is managed in Excel spreadsheets, and at any point in time, no one can accurately give any information, for example, about a particular asset, its configuration, applications running in it and versions. With IT audits getting stricter, some focus is returning to the asset management process, linkage to configuration management, CMDB (if there is one) and the actual IT operation mainstream process.
For most organizations, proactive spend management will require, at a minimum, an investment in time and resources to identify and track costs. Others will choose to implement an IT financial management tool to optimize cost management, reporting and reduction. Regardless, the objective must be to capture, aggregate and allocate costs to a cost model that enables financial analysis and spend tracking beyond the common chart of account structure of most ERPs or G/Ls.
One of the key milestones of an I&O evolutionary journey toward efficiency is alignment to the business. The ability to speak the language of business (finance), or to communicate the cost and value of IT investments, is critical. As a starting point, all organizations should develop a basic understand of what it costs to provide "IT services." This may be as simple as building a price list of IT goods and services that establishes a cost per line item that can be used to populate a basic "bill of IT," or price sheet. Alternatively, when more-accurate data and dynamic analysis are required, a robust IT financial management solution may be required. Either way, the solution selected should match the appetite and mandate for more-aggressive cost management. For more information on cost reduction via better IT financial management.
The cost-saving potential of process definition and task automation are significant given the redundant and manually intensive nature of IT tasks like provisioning, changing, deploying and approval cycles. Process automation is not new, but is often misunderstood and underutilized. I&O leaders must recognize that to be more efficient and reduce the costs inherent in IT operations, more-common tasks need to be taken out of the hands of individual task owners and crafted into defined best practices with less human intervention and more automation.
Automation of IT processes does not need to be overly complex, nor does it always necessitate significant new technology investments. Many management applications, across an array of functional silos, include the ability to define and automate processes. In the data center, for example, the definition and automation of processes is diverse and done with many management tools. These should be leveraged to define and implement time- and cost-saving processes and workflows. The addition of IT process automation tools, in turn, provides an orchestration layer for unifying functional workflows across different systems (for example, servers or routers/switches) into higher-value, business-oriented processes.
Cost reductions come as I&O leaders focus on automating tasks that include high levels of complexity, manual work, human intervention and oversight. Consider the potential savings of converting highly manual, approval-intensive, multitouch activities like equipment requests and purchases into an automated fulfillment model. In this sample scenario, the time and labor from purchase request to fulfillment includes five hours of active management at a labor rate that averages $30 per hour. A single instance can cost upward of $150 per request. In an environment where 100 requests are processed per month, the labor cost can run in excess of $150,000 per year. When automation can reduce the required human interactions by 30%, $50,000 savings per year, per purchase instance is a reasonable outcome.