Print

Comment

Email

Digg

Del.icio.us

Reddit

Twitter

Several challenges are faced by enterprises when it comes to designing data centers in an infrastructure. However, the use of a holistic design approach, best practices in power and cooling, and a proper metric system in place could help enterprises achieve significant cost benefits. The need for solutions with lower cost of ownership remains the topmost priority for CIOs, especially as demand for greater capacity and flexibility continue to increase in data center environment. Optimizing data centers has not only become a corporate social responsibility towards the environment, but has also become more of a cost saving measure. Concerns over infrastructure creation, power utilization, rising costs of energy, and the overall efficiency of data centers have risen. Moving on these lines, Dataquest in association with Sun Microsystems, Siemens and Emerson recently organized a tri-city event Optimizing Data CentersA Business Metrics Approach in Delhi, Mumbai, and Bengaluru.

(L to R): Furokh Santoke of Serco; Suresh Shanmugam of Mahindra & Mahindra Financial Services; Rajneesh De (moderator) of Dataquest; Anil Shankar of Shoppers Stop; and Bhushan Akerkar of TCS concluded in the Mumbai discussion that maintenance and uptime are the key to manage IT services either inhouse or when outsourced to third parties

Key Challenges
On behalf of Emerson, U Gopi, country manager, Data Center Business; and Ravindra Bapat, director, Data Center Business gave presentations on the challenges for designing data centers. According to Emerson, some of the biggest issues that IT managers face includeheat capacity, space constraint, power density, and data storage. However, the fact remains that there is no universal metric for data center output. In the absence of this metric, discussing data center efficiency may not be that meaningful.

(L to R): Francis Rajan of Bengaluru International Airport; Saumya Upadhyaya of Frost & Sullivan; Subbarao Hegde of GMR, Sudesh Prasad (moderator) of Dataquest; Prashant Pereira of Sify Hosting; and Mohan Ramaih of Wipro Infotech focused on the need to adopt intelligent management solutions during the Bengaluru discussion

Agreeing to this viewpoint, Neeladri Bose, director, infrastructure consulting, professional services, who represented Sun Microsystems in all three cities, said that metrics are key to achieve fundamental business goals, such as efficiency, risk reduction, and agility. He concluded by saying that there has to be a convergence of IT and facilities within an organization for a better design, functioning, and effectiveness of a data center. Further, Bose opined that virtualization and consolidation is the only way out.

Ravindra Bapat of Emerson expounded on the challenges in designing data centers	Neeladri Bose of Sun Microsystems felt that IT and facilities within an organization need to converge for a better design, functioning and effectiveness of a data center	UR Gopi of Emerson concluded that heat capacity, space constraints, power density and data storage are the biggest challenges that IT managers face while managing data centers

From Siemens, the event was represented by Gaurav Kumar, GM, marketing; and N Prasad, GM, marketing. In its brief presentation, Siemens discussed about the need for smart and safe data centers. According to both Kumar and Prasad, innovative products of sustainable value are needed to build smart and safe data centers in order to be environmentally responsible. They concluded by saying that just having excellent products does not mean having an excellent infrastructure. The need of the hour is to build environmentally friendly products.

(In Delhi) Vishwajeet Singh of FCM Travel; Praveen Khandelwal of Airports Authority of India; Rajneesh De (moderator) of Dataquest; Upal Chakravorty of DLF; and PJ Koul of Air India stressed on optimizing costs in designing data centers. This would save money as well as enhance operational excellence

The Whos Who
The key attractions of the tri-city event were panel discussions that were conducted by Rajneesh De, associate editor, Dataquest, in Delhi and Mumbai; and Sudesh Prasad, associate editor, Dataquest, in Bengaluru. The discussions highlighted on the challenges and benefits associated with data center optimization, while focusing on aspects like building the right infrastructure, adopting best practices in cooling and power management, optimizing storage and server capacities, and maintenance and manageability. In Delhi, the panel was comprised renowned CIOs and IT heads like Upal Chakravorty, CIO, DLF Universal; Praveen Khandelwal, DGM, IT Airports Authority of India; PJ Koul, DGM, IT, Air India; and Vishwajeet Singh, national head, IT, FCM Travel Solutions. In Mumbai, the panel discussion was joined by Suresh Shanmugam, national head, business IT solutions, Mahindra & Mahindra Financial Services; Anil Shankar, customer care associate & GM, technology infrastructure, Shoppers Stop; Bhushan Akerkar from TCS; and Furokh Santoke, associate vice president, Serco. In Bengaluru the discussion was powered by Francis Rajan, head, ICT, Bengaluru International Airport; Saumya Upadhayaya, analyst, ICT practice, Frost & Sullivan; Subbarao Hegde, CTO, GMR; Prashant Pereira, manager, products, Sify Hosting; and Mohan Ramaih, head, practices and technical support, Wipro Infotech.

Gaurav Kumar of Siemens stressed on the need for smart and safe data centers

In-house Vs Outsourced
The discussion in Delhi and Mumbai revolved around the advantages and disadvantages of self-owned data centers as against those of outsourced ones, and how to get an optimized solution meeting organizational requirements and business metrics. While DLF has outsourced its entire data center infrastructure to IBM and is relieved of any kind of operational and management issues, cost optimization is one key area where it needs to pay attention. In order to attain cost benefits, we have to keep ourselves updated as its all about the availability of latest technologies and how they can help our business in the most optimized manner. One more important aspect that we have to track is security, since we are letting access to our valuable data by a third-party. Third vital area to track is the server response time, said Upal Chakravorty from DLF. According to him SLAs are important as they provide assurance of all these factors, Therefore, SLAs are very crucial and they should be designed keeping in mind all the organizational requirements, he concluded.

FCM, on the other hand, has partially shifted its data center application to the hosted environment, while the rest is taken care by its global data center. Data center set-up itself is very complex. And managing the infrastructure is a very challenging job. Another challenge is the RoIhow to get best results out of the available infrastructure in the shortest possible time, explained Vishwajeet Singh from FCM, who supports the partial outsourcing model that helps less headache as well as better control. Both AAI and Air India have their own data center set-ups that are quite big and handle business that has huge networks spread across the country. Maintaining such a huge network is even more difficult as the operational and business load requirement keeps on varying. One key challenge for Khandelwal of AAI is how to keep the IT infrastructure up and running to meet that demand. To do that, it is essential to keep the infrastructure flexible and scalable. This is one vital aspect that we had to keep in mind while planning our data center set-up as well as optimizing the same, he said. Air India has, of late, consolidated its data center as a result of which it has successfully enhanced its overall business operation. According to Koul of Air India, it is very important to identify core business requirements and design a solution accordingly. The idea should be cost optimization, and not just reduction, that will not only save money but enhance operational excellence. The Air India data center is built on IBM mainframe technology. We wanted to build a system which could be consolidated anytime for optimum result and was scalable. Also, we wanted our data center to be flexible, scalable and able to support any kind of application as per our requirement, he explained.

According to Shanmugam of Mahindra & Mahindra Financial Services, dependency on data center has increased many times. Whether in-house or outsourced, maintenance is the key factor that has to be considered by every organization as uptime is of utmost importance. As per Santoke of Serco, outsourcing is a well-known tool for cost reduction. The pay-back period is also lower in outsourced model compared to in-house. However, in both the cases, the data center has to have the capability to be scalable and flexible, since business requirements keeps on changing, he opined. Explaining on the nitty-gritty of data center optimization Akerkar of TCS suggested to first identify how and why to optimize, It is important to identify how much of the data center is required to be optimized, and secondly, whether the implemented technology is scalable. He too agreed on the fact that outsourcing is a hugely cost-saving option as well as reduces manageability hazards. More and more people are now adapting to the outsourced model. It saves costs for up to 70% and requires no manpower for managing the data center set-up. One can concentrate on his core business area instead, he said. Shankar who is responsible for data center restructuring in Shoppers Stop, which has now outsourced its set-up, too supported Akerkars views and said, Its difficult to build the required skill-sets and expertise to establish and manage a data center for businesses like ours. In such scenarios, it is always effective to outsource. Commenting on data center optimization, he too agreed that its important to make the technology scalable and flexible to meet future business requirements.

The Way Forward
Addressing the issues on building a right infrastructure, Rajan of Bengaluru International Airport commented that the current rates of servers in the infrastructure may not be able to facilitate virtualization and therefore, there is a need for a specific kind of solution that would fit according to the organizations requirements. He added that going forward, virtualization and cloud computing will be key for any enterprise using data centers. Agreeing to that Upadhayaya of Frost & Sullivan said, In the next five years, the space requirement would grow to 30-45%. Challenges for enterprises are to cater to such a growth and that significant changes are needed by the enterprises to build their infrastructure. Consolidation is the key word, while there is also a clear need to plan about equipment purchase, its installation and commissioning. According to Hegde of GMR, a stakeholder structure needs to be in place. As a part of its strategy, GMR has set up value for money cross-functional teams (CFT) that gauge RoI and long-term investments from green strategies of the organization. However, Pereira of Sify Hosting had a different viewpointto develop the right infrastructure for any enterprise. Key to any business is optimization. Server farms should be created in a way that they can be scaled up as per the business needs. Everything can be deployed virtually, and running the system at an appropriate temperature is most essential, he explained. According to Ramaih of Wipro Infotech, Enterprises need to adopt intelligent management solutions. There is a need for high density racks, since computing power would double in the next few years. Virtualization as a key technology trend can also be looked as an alternative solution.

Piyali Guha
With input from PC Suraj
piyalig@cybermedia.co.in

Page(s)   1