Deriving Value from DCIM

We all remember the close association of Business Process Re-engineering (BPR) and Enterprise Resource Planning (ERP).

ERP had to deliver on a set of business outcomes defined from a BPR exercise, like improving inventory turns by 50%, or improving order to cash cycle by 20 days which would have direct impact on profitability.

As data centers are becoming increasingly capital intensive and operating costs of running them are mounting, we are seeing CIOs conducting data center audits.

In such a scenario, CIOs can leverage an emerging technology-Data Center Infrastructure Management (DCIM) software, which has been borrowed from the ERP world and can be effective in managing the opex and capex of a company.

Here are 2 examples, which we will examine in further details:

1 Cost Reduction: If a company is embarking on an organization-wide cost reduction program that also impacts the data center and it sets targets to reduce the capital outlays and operating expenses by a certain percentage over the next 3 years, DCIM software can help identify potential savings through:

• Better asset utilization
• Better floor and rack space usage
• Better power management

2 Higher Availability: If the data center is incurring penalties in not meeting up-time requirements as per (internal or external) customers' SLAs, then DCIM software can help through following ways:

• Identify potential single point of failures across the entire chain from application to power source
• Prevent failures through better alert management
• Provide ability to simulate Move-Add-Change to detect any potential adverse impact
• Provide predictive analytics for failure prevention before they actually happen

 

How to Reduce Data Center Capex and Opex

Reducing Site Infrastructure Capex: Building a data center is an extremely capital intensive project-it can be as high as $1,000 per square foot for a tier-3 data center.

Hence it is expected to have an extended life before it goes for an upgrade or expansion because of additional capacity requirements. Almost all data centers are either under-provisioned or over-provisioned in terms of floor space, power, and cooling capacities. This is primarily because of the unavailability of real-time data on space, power, and cooling requirements and the inability to simulate the exact need before the build project is kick-started.

Using real-time data, historical records and ‘what-if' scenarios, DCIM software can accurately determine the required site infrastructure capacity, thereby preventing over-provisioning and saving wasteful capital expenditure.

Reducing IT Capex: As per McKinsey, the average daily server utilization, across data centers as a whole, generally tops out at a low of 6%, creating tremendous waste in terms of the capital employed and energy used.

Since most data centers do not have this visibility, new servers are constantly being procured even when there is huge amount of available but hidden compute capacity in the data center's existing server base.

DCIM can discover this hidden compute capacity in a data center through real-time monitoring of server CPU utilization and can identify candidates for virtualization, consolidation, and re-purpose; thereby saving significant capital expenditure not only on server hardware but also on associated storage, networking, and software licenses.

Reducing Energy Costs: In order to reduce energy costs for a data center, it is first necessary to baseline today's energy use by the data center facility and its various IT and site infrastructure components and then identify the possible avenues to optimize energy and reduce cost.

Reducing Other Operating Expenses: Data centers employ a sizable workforce for monitoring and maintenance. In addition to this, a significant portion of the opex budget is allotted for support and maintenance of old hardware.

 

DCIM increases the operational efficiency of the data center by automating critical processes like asset management, remote ‘hands and eyes' support, capacity planning and provisioning, and thereby minimizes human resource requirement and hence reduces operating costs.

Reducing Cost of Downtime: Today's data centers are expected to maintain up-time in the range of 99.99%. In spite of building redundancies across the data center ecosystem, outages do occur, at an average of 2.5 times annually and the average outage duration is 134 minutes, as per Emerson Network Power. DCIM complements specialized Business Continuity Planning (BCP) software in multiple ways:

• Monitors every equipment in the data center, including the redundant components, at real-time and generates proactive alarms on any abnormal condition such as asset heart-beat, utilization, power consumption, environmental parameters, and physical security threats. The alarms can be sent to designated staff to initiate proactive action and prevent a potential failure.
• Provides an alarm dashboard with a summary view of all alarm conditions in the data center and highlights those that need immediate attention.
• Maps critical relationships among data center assets, application, and user groups and helps in doing impact analysis of an equipment failure.
• No wonder, CFOs are concerned about the ever-increasing capital outlays, operating (including power-related) costs and then building new data centers as space runs out. Time has come to arrest this. DCIM, as planning and management software, can help CIOs and CFOs to achieve this.