'Our Networks are the nerve center of our operations'

Comment

Digg

Del.icio.us

Scope International is the Global Shared Services Centre of the Standard Chartered Group. The company was incorporated in 2000 and commenced actual operations in Chennai from 2001. In a short span, Scope has grown to become an organization with over 5000 employees and constitutes about 8% of Standard Chartered Group's global workforce. It services all 56 countries of the Group and processes over 80 mn transactions a year in banking operations alone. Scope handles a wide range of value-added and complex services from banking operations to supporting global HR processes, finance and accounting services, software development and maintenance, and providing IT service and helpdesk support to the group globally. With these processes in the back drop, the bank's connectivity and network infrastructure assumes a mission critical role. Sarabjit Anand, head ITSC (W) and Country Technology Manager India and GSSC, Scope International, walks through the company's networks in an interview to Voice & Data. Excerpts

How critical are the Networks for you?
A few years back we decided to move towards a network centric model as part of our strategic vision. The initiative was started three years back, and we moved our core and critical application to our data center. The network centric model called for strategic and well defined networks for seamless operations. For instance, from a group perspective, we established 10 core sites spread across countries like India, Hong Kong, Singapore, Malaysia, UK, UAE among others. With this model in the backdrop, our network needs to be up and running always to meet the stringent requirements of Standard Chartered Bank's (SCB) operations. So in that perspective, our networks play indeed a mission critical role as all our apps sit in the centralized data center.

How is the model defined?
Our networks are defined in three layers. For instance, the first layer is the core, and then we come to the second layer that is distribution. The third one is the access layer. The ten sites that form the core are fully redundant MPLS based network. Then we come to distribution and when we take one of the core sites like India- within it we have two sites- Chennai and Mumbai, these are next level critical sites that drive the banking operation at the country level. So all bandwidth lands in these critical sites and then gets distributed across through the access layer.

How do you manage connectivity in your network model?
With ten core sites, the need for a reliable yet agile connectivity backbone is a must. We used different types of connectivity and multiple vendors, and arrived at a connectivity paradigm that meets our needs and requirements. International connectivity is primarily MPLS and we use IPLC in a small way. Our domestic connectivity is through leased lines enabling point to multi-point connectivity. We also use VSAT, but in a very limited way.

Choosing a connectivity vendor is critical for operations, given that what's your approach here?
Our connectivity infrastructure is made up of multiple vendors. We deliberately adopted that approach to arrive at a network that gives us resilience. Moreover, we are not tied with one vendor as multi-vendor approach gives us flexibility and heterogeneity in-terms of managing our networks effectively. We have defined standards in every areas of IT and network is no exception. For instance, we have something called 'standard built', which is configuring desktops or networks for our specific requirements.

When we factor typical banking operations such as the ones spread across branches, ATMs, e-Banking etc. How big are these challenges from a network perspective?
SCB is in 21 cities and 81 sites in India. Some of the sites are support offices. But providing connectivity is indeed a primary requirement for all branches to successfully connect and interconnect with each other. We have 198 ATMs in India and this calls for systems to always linked and hence network availability is extremely critical. Hence, we have a primary set-up and a back-up,with dial-ups as the back up when the leased line fails in rare cases. This combination of leased, dial-up and VSATs combination works well for us to arrive at securing peak uptimes for our ATM operations.

What about network security?
Security is of paramount importance. Here, there are two aspects to it-one is our own banking operation that's internal. The other one is external customers linked to us through our websites. In terms of managing the network security internally, we have well defined controls and access from inside to outside. For instance, only authorized people are able to access applications, and it is multi-tiered with various levels of access. We also have three-tiered firewall for Internet banking operations and the best pro-active practices to control spam and spy ware.

In terms of applications, can you give us a snapshot?
We have a core banking application called 'Hogan', which India uses and it is housed in our data center in Hong Kong. All our locations in India use this CBS application. We also use payment-based applications like RTGS with our member banks.

With ten core sites, the need for a reliable yet agile connectivity backbone is a must. We used different types of connectivity and multiple vendors, and arrived at a connectivity paradigm that meets our needs and requirements

What are your key goals through the network centric model?
Our two key goals here are availability and scalability. I think these are the two most critical factors that define the efficiency of the networks. For instance, the networks we have created should be available to the banking operations spread across many locations. From accessing the apps to a simple function like email, the networks form the backbone and any downtime in that is totally not acceptable. Meanwhile, any network has to be scaleable to the business expansion and growth coming out of it. For instance, SCB is going on an aggressive growth mode and our networks have to meet that without any stress. To take stock of things from a network perspective, a while back, we have launched a network re-engineering program aimed at further improving our overall network backbone. We are enabling our network to future demands. For instance, once the re-engineering program gets completed by this April, we would have bandwidth capacity that would be 5 to 20 times higher than what we are currently having.

On new technologies...
Since we adopt a standardized approach, any new technology has to be tested fully and we need to be comfortable to deploy that on an organizational wide basis. Hence, our use of technologies like Wi-Fi is limited to conference rooms. And here also, we have adopted restricted wireless usage through token-based access and not everybody can access it. As for technologies, we are looking closely at VoIP, where in we have created some proof points and exploring the possibilities of deploying it.

What are the key challenges CIOs face today in managing their enterprise networking infrastructures?
I think the twin challenges revolve around availability and scalability. Any network planning has to be done with growth in mind. Here you need to create a network that is able to take up the current requirement and provide round the clock availability. But QoS changes very frequently, and one needs to constantly scale up the networks to accommodate new requirements. For instance, applications and access patterns keep changing. Yet another key challenge I consider lies in prioritization of the networks. One should be always aware of what goes on in the network and need to put in place pro-active monitoring mechanisms in place that ensures a secured and reliable network backbone.

Shrikanth G
shrikanthg@cybermedia.co.in