Outsourcing network management and control system requirements enables enterprises to grow and focus on core businesses. That's the reason major chunk of large enterprises prefer outsourcing their Network Management Services (NMS)

NMS ensures proactive monitoring and management of various components of the network infrastructure that form the communication backbone for the organization. The typical functionality that network management generally covers includes fault management, configuration management, accounting management, performance management, and security management. These services can either be performed by in-house IT departments or outsourced to 'managed service providers'. The most essential technology that network management needs to adopt today is the 'Service Impact Analysis' area. This means that network management technology has to be capable of monitoring faults that impact 'business services' (e-mail, payroll, ERP etc. ) and depict the same accordingly.

With changes in network topology and increasing importance on value added services, the expectations from a network management application and services has increased multi-fold. The path that NMS is evolving into, include the development and usage of re-usable components to have a uniform platform to build applications on a single point configuration and management operation along with a hierarchical view to retrieve different levels of information. Additionally management system correlation, web based network management, increased focus on quality of service (QoS), and performance management are the other areas to be focused upon.

	EXPERTS PANEL
	A Prasad Babu, SE manager (India and SAARC), Juniper Networks Deepak Jain, GM-managed IT services, Wipro Infotech Kiran Bhagwanani, VP-sales (APAC), HCL Comnet Nagamani Murthy, VP (voice and next generation networks), Wipro Technologies Rajendra Dhavale, consulting director (India and SAARC), CA Ranajoy Punja, VP-marketing (India and SAARC), Cisco Systems

With the growth in applications, there is also a need for achieving greater bandwidth efficiency, which impacts the cost of network ownership over time.

Some of the technologies that enterprises would move towards with a view to achieving the above are AI based event co-relation, cross domain co-relation, root cause analysis, route analytics, network performance management using specialized tools, network configuration management, VoIP service assurance, storage management, VPN management, and application response time management.

Converged networks (IP networks that carry more than just data, they also carry voice and video information) are on the rise today. Organizations look to leverage their existing infrastructure and investment by incorporating new technologies such as voice and video transfers on existing networks.

The key to successfully implementing converged networks and mobile access solutions in the enterprise is to manage, monitor, and secure entire infrastructure (including end-devices) such that the impact of component failures or performance degradation on 'business services' can be depicted.

Also we need to get away from the myth that VoIP alone is convergence. Convergence in an enterprise needs to happen at the application level where we have unified applications across different network domains. In a service provider scenario, convergence needs to happen at the access layer where multiservice access nodes and multiservice engines are able to aggregate services and traffic.

ENTERPRISE CHALLENGES
Apart from the challenges that enterprises typically face from competitors and new business models dictated by the market, three key business challenges that CIO's face today include reduce cost, manage risk, and ensuring infrastructure is always 'ON'.

Another challenge that CIOs have accepted today is that the existing tactical tool sets that discretely manage pieces of technology (networks, systems, databases, applications) and do not integrate with each other, are simply not providing value.

Most of the enterprises are today moving away from a network-only view of network management to a more holistic approach that encompasses looking at overall application availability and end-user experience. To this end, the focus is not just on SLA, which of course is the start point, but also on performance management, event co-relation, and application response time.

With the growth in applications, there is also a need for achieving greater bandwidth efficiency, which impacts the cost of network ownership over time. For this, what is needed is deep visibility into network utilization both at the application and end-user level so that bandwidth can be optimally managed rather than over-provided.

The key challenges for enterprise comes in the form of monitoring the network for performance when the networks are becoming more and more complex, controlling and operating at a high QoS for a prolonged period of time, adding innovative and rich value added services in complex networks for additional revenue streams and of course mobility management.

Despite the challenges, network managers operate in today's market with high degree of reliability and QoS. To get the opportunities converted into right kind of response enterprises needs to have edge over their competitors by providing innovative value added service, proactive customer management, sophisticated performance monitoring system, and decrease the time-to-market the solution.

SOLUTIONS
Managed Services is a boon for lot of organizations that decide to focus on their business to start with, rather than the upkeep of infrastructure. With appropriate SLAs in place that ensure uptime and response time for network infrastructure; organizations let the managed service provider take over the management. In certain cases, this has definitely contributed to reduction in TCO where the cost of network maintenance was higher than what is being paid to the MSP.

There is another school of thought-'Why do I outsource my infrastructure management, especially when it is such a crucial component of my business?' Such organizations generally develop in-house expertise and manage the infrastructure themselves. ICT departments in such organizations are more of value centers rather than cost centers. They actually provide business services to business units based on the existing IT infrastructure with appropriate SLAs in place.

It can only be answered by three key priority areas that most organizations are aware of-people, process, and technology.

Enterprises face a huge challenge, while trying to reduce the total cost of ownership and keeping the QoS consistent. A sophisticated network management system can help in reducing the operation cost and minimize opportunity cost. Operational cost can be greatly reduced by automating network management applications, bringing in more automated intelligence in network operation, and centralized NMS. This will also result in fault tolerance to the network through correlation, thus, bringing down the opportunity cost.

RISKS INVOLVED
The risk involved from the customer end is choose the right experienced partner having a mature transition model and also having the ability to manage multi-technology and multi-vendor networks.

Outsourcing the NMS of a company does bring in an element of risk that includes security concerns, loss of control, and vendor lock-in. The primary and predominant risk involved when outsourcing the network management of the company revolve around the capability of the service provider to access privy information of the company by virtue of its control over the network This risk can be reduced by having adequate SLAs in place and also having a trust worthy and established vendor as outsourcing partner.

DEPLOYMENT TRENDS
The prime challenges that face wireless networks today include security and technology, interoperability/standardization (802.11). Although wireless networks (Wireless PAN–via Bluetooth/LAN via Access Points/ WAN via GSM and GPRS) offer flexible connectivity to mobile users. Security has become a prime concern for organizations implementing the same. Some major concerns that organizations face include:

Detect Rogue Access Points-This is one of the biggest menace with wireless network unless it is securely implemented. This allows rogue users to log-on to corporate networks by setting up access point type apparatus to receive data signals
Bluetooth and 802.11 Interference-Both Bluetooth and 802.11 operate at the same frequency thus causing network problems when both are in use at the same time. This results in slowdown or dropping of the VPN connection
Protect Wired Networks-With the introduction of wireless access points and the security issues that come with it, it becomes imperative for organizations to boost security on the wired LAN where business critical applications are hosted
Overcome 802.11 WEP Security Holes-Wired Equivalent Privacy (WEP) provides data encryption based on RC-4 cipher and can be easily cracked. It is crucial to detect such devices and protect them using better authentication technology such as extensible authentication protocol.
Wireless Networks in an Enterprise is primarily on WLAN technologies. The WLAN being a shared network leads to three key challenges that include QoS, security, and management.

Ability to deliver QoS for applications over a wireless network is a key challenge. Wireless being a shared access network, needs significant planning and engineering before deployment. For security deploying wireless controller based network rather than access point based networks enable centralized security and administration and policy enforcement. Wireless networks have a wired component and a RF component. The network management tools should be able to discover and provide co-relation/root cause analysis and performance management to the wireless network. Specialized tools are available to deliver the same.

TECHNOLOGY TRENDS
There are multiple tools and technologies that allow organizations to gauge the performance of their networks. On one end of the spectrum are freeware, shareware, and open source tools that simply collect SNMP based data for charts depicting utilizations, error rates and availability etc. On the other hand, there are enterprise class tools that cater to measuring performance of every aspect of large networks (corporate networks that span continents and service provider networks etc).

As organizations move up the IT maturity model they start relying more and more on the IT infrastructure to move their business ahead. They need better performance management tools that help them pin-point the trouble areas quickly, thus reducing the mean-time-to-repair.

Some of the key technologies that can help in network performance management include.

Intelligent threshold management-Most tools provide comparison of performance data against a static threshold (Alert network administrator if link utilization goes beyond 70%). Today such comparisons have little value. The threshold itself is dynamic. It is abnormal for link utilization to increase beyond 70% during business hours. But this threshold violation could be perfectly valid for post-business hours where data is being replicated from various branches to the central location.
Better Analysis of performance data-Not too many network administrators have the time and knowledge to analyze performance data. They need to act fast in case of problems. Technology can come to the rescue with pre-analyzed reports that depicts the 'At-a-glance' status of the network, it helps resolve performance problems faster.
Ability to provide performance statistics for entire technology stack-It is not enough to just look at a network performance chart to diagnose what has gone wrong. It is imperative for implemented technology to provide an insight into the performance of other aspects of the infrastructure ( such as systems, databases, applications, web servers etc. ) to accurately diagnose the problem area.

Despite an assumption of the contrary, the enterprise network is not dependent on the network elements only. In spite of the fastest elements, the enterprise network can be hampered by slow data traffic. Therefore it is very important to design and plan networks to match the speed of business. To accomplish this, network planning tools have to be used for planning and enhancements. Constant monitoring is imperative to pre-empt and also solve bottlenecks and potential failures. In addition to this, prioritization of traffic, load balancing, increasing bandwidth, and minimization of communication between network elements are some of the additional measures that ought to be taken to improve network performance.

Also technologies such as content networking, application oriented networking, VPN acceleration, QoS, traffic shaping, and policy management platforms would also help determining improved network performances. However, deployment of market intelligent applications in its peak market window still holds the key for quick returns on the investments made.

The point to note here is-just procuring the tools will not ensure true RoI. It is about correct implementation and sustenance of these solutions that allows organizations to realize value of investments.

RECOMMENDATIONS
A network has little value, if it cannot be managed properly. Some of the best practices that network managers can adapt would be using Automation whenever applicable, have a graphical view of the network being managed and use good network management tools to predict and pre-empt potential problems such as network imbalance, data influx etc. Some important function includes:

• Factor in redundancies at the design stage, especially for critical components.

• Focus on overall application availability and not just network availability.

• Manage, rather than over-provide, bandwidth.

• Rely on the experience of specialized Managed Services Providers (MSPs) for a holistic, integrated approach to application availability that will lead to enhanced end-user experience.

Network managers can move up the value chain by implementing the right network technology that provides business value and ensure that the same is available and performing at all times. Also network managers are best advised to implement
(or out-source) management technology that helps them automate functions such as discovery, fault management, root cause analysis, reporting, and performance bottleneck detections. This ensures that they spend quality time on innovative job functions such as network re-architecture for optimal usage, implementation of cutting-edge technology such as MPLS.

Rahul Gupta
rahulg@cybermedia.co.in