Tata Tele revamps DC infrastructure to meet its rising demand

Tata Teleservices (TTSL), one of the leaders in the Indian telecom space, was growing leaps and bounds when it realized the need to upgrade its IT infrastructure to sustain high demands in the future.

TTSL took a strategic decision to consolidate its backend infrastructure facilities in Hyderabad. As a result, a new data center, technology area, and user facility, together known as Gyanpeeth, 3.4 lakh sq ft in area, was built and developed from scratch.

The earlier data center, also in Hyderabad, was not fully geared for this growth and hence, the need to migrate was contemplated.

TTSL has its large central primary data center in Hyderabad and site level DR center in Delhi and there was an urgent need to migrate the business critical primary data center into the new facility in Hyderabad within 6 months as the building lease of old data center facility was expiring by the end of March 2012.

At the same time, TTSL was working on consolidating its centralized operations in all spheres-IT, technology, network, NOC, etc-at its own new and state-of-the-art facility in Shamshabad, a quieter suburb of Hyderabad.

A Logical Step

Hence, it was a logical move to migrate the data center to this facility, which offered world-class amenities and almost limitless potential for growth and expansion. One of the important considerations for the company in the migration was the opportunity; it offered to upgrade their deployment architecture.

TCS being the IT Outsourcing partner executed the large data center and user migration program. TIA standards were adopted to build tier-3 compliant new data center with state-of-the-art architecture.

TTSL has a long-term partnership with TCS as its strategic IT outsourcing partner. Since 2005, it has helped TTSL to manage its costs better and build a portfolio of about 80 applications serving 90mn subscribers. The applications were a combination of COTS and custom built applications. The approach was to build custom applications wherever possible to ensure cost effectiveness and maximum value creation.

Due to the prevalent ‘don't fix if it ain't broke' mindset-particularly for mission critical systems, many CIOs are constrained to live with a build-as-it-grows design which does not always result in elegance and efficiency.

The Challenge

The biggest challenge it faced was to take an aggressive target of 6 months time-frame for completion of mammoth data center migration with optimal cost and without impacting ongoing IT operations. The idea from TCS was to improve standards of overall IT efficiency through modernization, optimizations, and innovative solutions.

Migrating several thousand servers, switches, routers, and storage equipment, which host several petabytes of data and hundreds of mission critical applications, without experiencing a single moment of business discontinuity was almost an impossible task set by TCS.

Shifting production stacks turn-by-turn to an interim staging environment, which was not always a replica of the original, came with its own risks that required sound mitigation. Data replication between the old and new sites was a critical hurdle, due to the sheer size and need for speed. For the first time in the world, a 10Gbps link was commissioned successfully for this.

Getting multiple partners and agencies to collaborate to work as one unit with a single goal was imagined to be a daunting task but proved to be one of the big successes. Finally, as with all projects, there were tough timelines and a need to adjust their roll-out was an added challenge.

What it Did

1 The company identified the need to establish a dedicated LAN connectivity between old and new data center to boost the time taken for exchange of data during migration and also ongoing live. They ensured to implement the best possible industry standards CAT6a cables in the data center to ensure there are no latency issues at peak load.

2 It re-designed the passive components to ensure better cooling and power facilities apart from improved network cables within the DC.

3 The organization believed the need to virtualize the entire server farm to increase capacity utilization of existing servers.

4 It upgraded most of their applications and databases to support tighter integration of their work flow and processes. Since the infrastructure was completely heterogeneous, TTSL had to carefully integrate the changes in the application and the databases which could support consolidated management for the same.

5 The overall networking was revamped with improved switching and routing solutions with firewall architecture to make optimal use of data flow and security protocol.

Benefits

The improved deployment architecture has clearly provided an ability to isolate and fix problems much faster leading to better business performance. TTSL focus on using the migration as an opportunity to make their architecture more efficient also led to optimization in capex, thus providing long-term cost savings.

TTSL identified several servers and stacks that they could either sun-set or consolidate, thus creating a leaner design. By keeping an eye on power consumption and cooling requirements, TTSL achieved substantial overall savings in opex as well, while contributing to a greener environment. An overall sense of exhilaration prevails at the end of a job well done, which is also a reward in itself.