|
Peer-to-Peer
Upgrading network for improving efficiency
Through the deployment of Cisco solution, Irevna has been
able to address the issues related to the availability of Internet, mail service,
and VPN tunnels, writes Nivedan Prakash
Irevna
is the high-end financial research and analytics outsourcing division of Crisil,
one of Indias leading ratings, research, and risk and policy advisory
company. The companys clients include financial institutions with investment
banking operations, insurance companies, asset management firms and the corporate
sector.
For clients, spread across the globe, Irevna is an ideal knowledge partner in
cutting-edge financial analytics and thought leadership. The companys
domain expertise, rapid execution, and quality output helps it maintain deep
and engaging relationships with clients.
Business challenges
Irevna is a technology-savvy enterprise and uses sophisticated IT solutions.
But to help meet its clients distinct business needs, it required high
Internet availability. The company has offices and delivery centers at multiple
locations where Internet connection is provided by different Internet Service
Providers (ISPs).
However, this made it difficult to ensure high Internet and mail server availability
as well as VPN tunnels uptime. In fact, Irevna already had a Border Gateway
Protocol (BGP)-based network architecture that used a Cisco solution to ensure
high Internet availability. Nonetheless, it wanted to enable Performance Routing
(PfR) across locations and this involved routing Internet traffic along the
best performing ISP link across locations.
To address these issues, Irevna went ahead with the Cisco solution because of
Internet performance routing, scalability and growth path with respect to latest
technology ahead of curve and better service delivery, high availability of
VPN, and load balancing of multiple ISPs that are geographically distributed
across the country.
Commenting further, Natarajan Kumarasamy, Head-IT Infrastructure,
Crisil Group, said, Ciscos ability to deliver a complete solution
fabric, encompassing the data center, corporate LAN, intranet and network edge
that addressed highly available, flexible, secure and scalable solution for
network infrastructure was the key driver for Irevna choosing it. The partners
ability to design and integrate this was also an important criterion.
Although other vendors were also evaluated for this solution, Ciscos ability
to deliver and complete PfR solution on top of BGP for all Internet based solutions
across different geography locations encouraged the company to choose its solution.
| Company |
Irevna, a division of Crisil Group |
| Solution |
Implementation of BGP-based WAN architecture and
Performance Routing technology from Cisco |
| Aim of implementation |
To enable high Internet and mail server availability
as well as VPN tunnels uptime |
| Cost of implementation |
Rs. 9 lakhs |
| Phases of the implementation |
2 Phases |
| Benefits |
- Able to use the network to handle any
business needs related to Internet availability
- Significant difference in terms of load distribution
and applications
- Able to offer much higher SLA and secure environment
to the customers and employees
|
Phases of implementation
|
"Cisco's
ability to deliver a complete solution fabric, encompassing the data center,
corporate LAN, intranet and network edge that addressed highly available,
flexible, secure and scalable solution for network infrastructure was
the key driver for Irevna to choose it"
- Natarajan Kumarasamy
Head-IT Infrastructure, Crisil Group
|
The implementation of the Cisco solution took place in two
phases. Phase I included BGP implementation across different geographies and
Phase II involved deployment of PfR over Irevna BGP Cloud.
As far as Phase I is concerned, high availability of Internet was one of the
core requirements for Irevna as a business need. Even though different locations
of Irevna used different ISPs, there were issues in addressing the availability
of Internet, mail service and VPN tunnels.
Considering the business needs, Cisco came up with the BGP-based
WAN architecture, which met Irevnas requirements. This involved using
Metro Ethernet links to form a ring which connects the WAN edge of all locations.
Steps to tackle the Internet redundancy were implemented in such a way that
it addresses ISP, hardware, and media failures.
An enhanced object tracking mechanism was also used with Hot Standby Routing
Protocol (HSRP) functionality to address the issues associated with the Ethernet-based
termination from ISP. Open Shortest Path First (OSPF) was then used as an Interior
Gateway Protocol (IGP) in the WAN edge across locations, and fully meshed iBGP
architecture was adopted across all Internet routers. The Cisco 2800 series
routers were used for this architecture.
Currently Irevna is in a comfortable position to handle the Internet-availability-related
business needs by achieving the automatic ISP failover across multiple locations,
as well as within a particular location, commented Kumarasamy.
For Phase II, though the company had BGP-based network architecture that uses
Cisco solution to address the Internet availability, it wanted to enable performance-based
routing PfR across all locations ISP and routing the Internet traffic along
the currently best-performing ISP links across locations. Cisco recommended
its best-breed PfR technology to meet the above requirement.
PfR architecture is well planned by Cisco Implementation Partner by keeping
Irevna BGP network architecture in mind. Irevna-Cisco WAN components have been
upgraded as per the Cisco recommendation to meet the PfR requirements,
added Kumarasamy.
Since Irevna BGP architecture involved multiple geographic locations and multiple
ISPs, the Cisco partner and the design team then analyzed Irevnas network
architecture to come up with a PfR design which included positioning the two
PfR componentsMaster Controller (MC) and Border Routers (BR).
The MC redundancy has been built to address MCs single point of failures
and its availability was addressed by using HSRP on edge routers. Apparently,
PfR policies were applied globally in the BGP network to enable all the Internet-based
applications to utilize the best-performing Internet path, based on defined
PfR parameters.
Kumarasamy further pointed out that apart from this, we identified business-critical
applications across locations and then applied unique policies based on combination
of different parameters such as delay, utilization, etc. After observing
PfR behavior in the BGP network, we came to know that there is a huge difference
from the plain BGP environment and the applications getting routed through best-performing
ISP path, in terms of load distribution, he added.
Meanwhile, Irevna made an investment of around Rs. 9 lakhs to build this solution.
And the only challenge that the company faced during the implementation of this
solution was that they were not able to deploy PfR using Fluke PfR Manager that
resulted in delaying the whole process of deployment.
Reaping the benefits
At a time when companies strive to lower the cost of running their networks
and increase the productivity of their end users with network applications,
Irevna was able to use the network to handle any business needs related to Internet
availability.
After observing the PfR behavior in the BGP network, Irevna saw a significant
difference from the plain BGP environment in terms of load distribution and
the applications that got routed through the best performing ISP path. As a
corollary, Irevna was also able to offer a much higher SLA and secure environment
to its customers and employees.
In addition, the solution deployed was able to better resource utilization,
transfer knowledge to their staff which enhanced skill-set, give them competitive
advantage that opened up new market opportunities and increased the stability
and availability of their network, highlighted Kumarasamy.
Additionally, Virtual Switching System (VSS) technology has been implemented
for the first time in LAN architecture, where all the modules would be in the
active state even in virtual architecture. The VSS deployment has been implemented
with the exclude interface methodology.
With this methodology, the critical servers would be retained access during
the dual-active detection. Furthermore, this VSS technology allows for new enhancements
in all areas of network design, including high availability, scalability, management
and maintenance, as well high availability for all critical servers are attained
through this deployment of VSS architecture.
nivedan.prakash@expressindia.com
|