Of late, Indian corporates have exhibited keenness in the adoption of WAN,
recognizing it as an efficient company-wide communication platform. Also, with
the realization that the Internet offers a cost-effective network communications
solution, they have significantly increased their Web presence. ERP networks too
have been set up.
With this has emerged the need to keep the network running ‘at all costs’.
This is the basis for organizations opting for outsourced network management,
which assures commitment levels through service level agreements (SLAs).
Network Management
Today, many businesses must operate their networks in a multi-technology,
multi-vendor environment that is increasingly costly and complex to run and
manage. As the core business functions rely increasingly on networks for
communications and delivery of critical data, network management strategy is
driven by the need to minimize network failure, improve service quality, reduce
operating costs, and enable new business opportunities.
Keeping the network running efficiently at all times, especially at the time
of a rapid expansion, requires putting together a team of network professionals
with necessary technical skills, establish processes and polices to control the
activity, and provide skillful management of the entire operation.
Managers responsible for network operations must consider the following:
- The most formidable challenge for organizations is to hire, train and
retain a cadre of highly skilled network professionals who can make sense of
the many-faceted elements of a modern enterprise network. A focus on these
goals may be more important to success than just a good technology - Processes and procedures that can maximize the value of the network
management technologies deployed, demand careful attention. As the
capabilities of network operations centers (NOCs) grow, adaptation of
staffing requirements and organizational adjustments are to be made, keeping
the future requirements in view - Outsourcing offers relief from staffing and skills dilemma, allowing
organizations to attain instant network expertise for maximizing network
performance. It also helps with critical functions, such as security
NSP Perspective
A few NSPs like Satyam Infoway and Hughes Telecom have been
maintaining network management centers (NMCs), which have also deployed
monitoring services for clients. While these have been restricted to
notifications about issues such as link downtimes and traffic observations, NSPs
see emerging benefits in taking up larger responsibilities of managing client
networks. Also, given the all-encompassing nature of the Internet, it will not
be a shock to read about network managers who are not located in the same
country.
A recent worldwide network industry survey of NOCs across the
globe by Lucent brought out the fact that managed service providers (MSPs) are
looking at highly integrated solutions, which cater to the complete needs of the
corporate world, right from remote desktop management to information security
management.
The delivery of managed network services usually requires the
creation of a NOC, where a core set of tools can be used to enable a relatively
small number of experts to efficiently perform remote monitoring and management.
A key factor in profit equation for a managed network service is the NOC’s
operational efficiency–the number of customers and services that each NOC
engineer can support simultaneously. The more the tools and related training can
be utilized across multiple customers, products, and services, the more cost
efficient the NOC operation becomes. This is partly due to reduced software
acquisition and maintenance costs, but even more due to the ability to assign
NOC personnel to tasks across customers and services as needed rather than
having narrowly trained specialist groups.
Types of Outsourcing
The type of outsourcing that a company can go for depends on
the criticality of systems, the amount of interference it is willing to allow
with its network, the amount of expertise available in-house, and the kind of
manpower that can be put to use on more profitable projects.
The following levels of outsourcing are prevalent:
Level I: Monitoring and alerting
At this level, the NSP is responsible for monitoring all
critical points or terminals of the network. Any fault detected is immediately
reported to the customer. In such an arrangement, the NSP corresponds with the
support staff on the client’s premises. The client doesn’t have to invest in
expensive monitoring software. The NSP does that and distributes the cost over
other similar projects. The client is required to maintain an IT support staff
that can act on alerts.
Level II: Monitoring and remote repairing
In this case too, the NSP monitors the network status of the
client. In addition, the service provider is also able to remotely rectify any
faults detected. The support staff at the client’s premises is required to
rectify faults when the service provider is not able to do the same.
Level III: Remote repairing and on-site service on call
The NSP commits to remote or on-site support. It remotely
monitors different client networks and also provides a support team, which
visits the client if remote repair is not enough. The client need not maintain
separate staff for monitoring network faults and can concentrate on more
advantageous tasks.
Level IV: On-campus presence as well as remote monitoring
This signifies the highest level of SLA. The NSP performs
remote monitoring and also has a person staying at the client’s premises. Such
arrangements have been observed in offshore network monitoring arrangements.
Pay as You Use
Tool-based infrastructure management entails heavy
investments in systems and tools, in addition to ongoing costs incurred on
manning the control center. By far the most alluring value proposition of an NSP
is the pay-as-you-use model, which obviates the need for organizations to cough
up huge upfront investments in management infrastructure. One way of looking at
an NSP service has often been an ASP type of service, which offers the
functionalities, delivered by the capital-intensive combination of tools and
platform on a rental basis.
Upward-bound costs of infrastructure management are largely
on account of the requirement of highly skilled manpower that is perennially in
short supply. Management of this skilled manpower in terms of retention and
maintenance is also proving to be a costly and troublesome task, which is best,
outsourced.
Table 1: Cost |
|
Capital Expense |
Operational Expense |
Hardware cost |
Hardware upgrade cost |
Cost of EMS tools |
Cost of upgrading EMS tools |
Cost of ancillary |
Cost of upgrading other software/tools |
Implementation |
Premises’ rental |
Facility upgradation |
Manpower cost of network management staff |
Cost of agents and |
Bandwidth/communication costs |
Costs of recruiting |
Department overheads |
Training of manpower recruited | Retention, and retraining of new staff |
Table 1 presents a summary of costs, both under the capital
as well as the operational head, which go into establishing and operating a
management infrastructure.
As can clearly be seen from the above tables and charts, a
substantial portion of costs incurred for in-house management of the network is
taken up by the investment in core tools required to run the network management
operations. Typically, an NSP is able to derive scale economies on this
investment, the absolute cost of which is fairly independent of the scale.
Economies also arise on hardware costs as the cost to power ratio for hardware
is non-linear. This essentially means that if larger hardware equipment can be
justified for the load, it is more economical on a per task basis.
In case of operational costs, manpower costs form more than
50% of the recurring costs. However this benefit is not as significant, on a
cost basis, as the benefits from investment cost reduction. This is primarily
because manpower costs are predominantly variable in nature. An NSP however is
able to offer a better value proposition here because he is in a position to
utilize this manpower more efficiently with lesser idle times. Over and above,
this is the release from the efforts and costs involved in locating and training
this manpower and retraining costs if they quit.
Benefits of Outsourcing
Break-up of |
|
Capital Cost |
Head Value |
Core tools and installation |
50-55% |
Agents and loading |
~25% |
Hardware |
15-20% |
Ancillary tools |
~5% |
Hiring and training | 2-3% |
NSPs having the expertise in the network management field can help corporates
extract the relevant information from a host of data collection computers, such
as network management appliances driving SNMP traffic, making them into useful
data. Such equipment is located at major points on the Internet backbone.
Depending on the resources available with the NSP, many such monitoring points
may be maintained in different countries. The devices, simulating real
end-users, can send http requests to various Web servers periodically, and
measure the turnaround Internet response time. The NSP through its network
resource center (NRC) can analyze the collected data and present it in a
user-friendly Web-based report. Such reports are available in real-time and
users can also look at historical information of the Internet response time.
NSPs’ unique advantage in server and systems monitoring allows customers to
monitor their mission-critical servers’ status, irrespective of which Internet
data center (IDC) they are located in, from major Internet backbones around the
world. While the IDC’s monitoring service does measure the server’s health,
it is limited to merely isolating server-related bottlenecks. Other important
measurements, such as end-user response time, which falls outside the IDC’s
measurement parameters, are not included.
The benefits of an NSP service go beyond 24x7 availability and protecting the
mission-critical e-business site with round-the-clock monitoring. They include
proactively solving site performance issues before users find them. Other
benefits, which the outsourcing firm can look forward to, are:
End-to-end Performance Measurement
Break-up of |
|
Operational Cost |
Head Value |
Manpower |
55% |
Core tools maintenance |
~20% |
Hardware maintenance |
10-12% |
Overheads |
10% |
Rentals |
4-5% |
Ancillary tools maintenance | 2-3% |
NSPs measure the server performance through two different mechanisms:
Internal server monitoring within the IDC, and end-user response time
measurement through network management appliances located at major Internet
backbones. NSPs therefore can identify the root cause of any performance
problems more efficiently.
Without having to invest in additional hardware, software or human resources,
an NSP allows the outsourcing firm to focus on its core business and leave the
task of performance management to the experts.
Problems Pinpointed
The NSP can pinpoint the root cause of a performance-related issue. By
following a systematic tracking process and correlating to performance metrics,
one can isolate performance problems to specifics, allowing the customer to
quickly zoom in on the issue for effective resolution.
Initially, it is logical to expect large corporates to reap the benefits of
strategic outsourcing and NSPs can expect projects for managing reasonably sized
WANs. But as NSPs mature and more awareness is created about the possibilities
in outsourced management, more firms, regardless of their size, will be found
outsourcing their critical operations.
Prof. B Sridhar is associate professor; Anand R, Kamalakkannan E, Koka
Dinesh Dutt, John Ben P, and Preeti Krishnan are students at IIM Lucknow