Top Ten Mission-Critical Network Management System Functions

Managing network changes has never been more critical. Here are ten functions a good network management system should provide.

Complex enterprise networks require new ways to manage the myriad technology changes we see every day. Simply reacting to change is no longer acceptable. Savvy IT organizations are taking a more proactive, automated approach to change management in order to balance speed with process. This approach ensures that rapid infrastructure change can be accommodated and digested while still delivering on business expectations.

Such change management has become an expensive endeavor in this era of multi-platform/tier architectures. To stay current with the network management systems (NMS), we offer this list of functional considerations for managing a network in today’s dynamic-infrastructure environment.

1. Network discovery should be automatic and easy

This critical point will ultimately affect all aspects of network management. The challenge is to discover assets quickly without affecting mission-critical business processes. For example, in the case of an investment bank, you need the real-time ability to discover your infrastructure during peak trading hours to capture changes. Discovery is more than just understanding the device type. Frequent and accurate connectivity data to complete your NMS solution is of prime importance. You can choose to eliminate solutions that require total network re-discovery every time something changes or those that don’t offer multiple methods for discovery (such as automatic, incremental, on-demand, and seed-file).

2. Database integration should be flexible

First and foremost, you need a database; otherwise you’ll completely lose all historical data and integration capabilities will be severely limited. If this condition is met, the database must be an industry standard; proprietary databases may be difficult to work with and may require costly professional services. Also important: a database with multiple data-export capabilities. This is crucial for companies looking to integrate and/or aggregate data. Finally, the data managed should provide the flexibility of an advanced format such as XML. In today’s environment, ASCII alone isn’t sufficient.

3. Leverage fault management features for accurate discovery and connectivity

Most traditional fault management tools are able to identify the root cause of an outage. However, proactive support teams need more information than just “an outage has occurred.” They need enriched event information, such as the impact of the outage: “What service will be interrupted from the customer perspective? What are the details of the root cause (not just packet corruption, but the type of packet corruption)?” This information is critical to effectively and quickly resolve issues. As the environment changes, the fault management and understanding of connectivity that feeds it, both need to dynamically change. All this data must be incorporated as part of a larger, automated chain of NMS processes in order to keep up with a changing infrastructure.

4. Manage performance in real time, near-real-time, and over time

Performance Management means different things to different support teams. Such perceptions can include low OSI level performance of WAN circuits, or a high OSI level GUI application reaction speed. Be sure to understand the performance criteria that are necessary for your environment. With accurate change-enabled data, look for performance tools that leverage Layer 2 through Layer 4 connectivity so NOC/operations staff will not only see a link performance degradation, but also note the services (such as Mail, Siebel or even Web services) that are affected by the degradation.

Be sure you can see historical as well as real-time performance statistics with ease; this is important when reviewing the performance of any part of your mission critical infrastructure. WAN performance statistics showing at least frame relay, ATM and leased line are mandatory. Also, ensure your solution can be adapted to new technologies without having to re-write or create new code. Other issues to consider include flexible performance-data rollup, data collection flexibility, full Web interface functionality, and a comprehensive set of standard report templates.

5. Asset/Inventory management should handle change

If you look for a solution that only handles asset/inventory management, you can expect high professional-services costs for deployment, customization, and (ultimately) integration with your remaining NMS tools. You will also be duplicating polling efforts—no vendor will surrender polling. It makes sense, nevertheless, to identify tools that add asset/inventory management out-of-the-box. After all, if a vendor is already communicating with a device that is managing fault and/or performance, it could as easily collect and manage asset information. With an accurate and quick discovery mechanism (see the first factor in this list), a vendor should be able to identify asset data and showcase it accordingly. When considering infrastructure, information for spare ports, serial numbers, names, locations, and version numbers is ready to be collected. More advanced tools should be able to report on these asset/inventory findings, including asset/inventory changes.

6. Use the right reports: stand-alone or integrated

As with all other aspects of NMS management, if you buy a stand-alone reporting tool, you’ll need to keep it integrated with other components. Today, you can find reporting tools that are logical additions of discovery (Factor 1), fault (Factor 3), performance (Factor 4) and asset/inventory management (Factor 5). Reporting tools must also be sufficiently flexible to include any type of data or change.

Likewise, you’ll want to manage report-scheduling frequency to provide real time or near real time availability. With a fully integrated solution, information such as Service Level Agreement (SLA) reports is possible. Such tight integration with reporting allows users to drill down from within SLA reports to see why SLAs aren’t met because of outages or interruptions in your network, system, or application. If you outsource your network management, , having the ability to see what’s going on can save thousands of dollars.

7. Use remote administration features

A Web interface is a requirement that refers to full user functionality as well as administration capabilities. Because we live in a virtual world, you may not always be in front of your NMS system when you need to perform an administrative task. Be sure your NMS solution, therefore, has the ability to automate tasks to minimize administration. If it is inevitable that you need to perform administrative tasks remotely, at least make sure you can move, add, delete, and change devices and their properties under management controls.

In addition, the Web interface should offer you the ability to check mission-critical processes and be flexible enough to co-exist with other applications on the server. From the user’s perspective, if you cannot fully operate each and every feature from a client interface (in other words, if you have to be on the server to see client features) you need to reconsider that tool or tools. Networks change, so do organizations. You need a solution that can be completely accessible and administered remotely.

8. Manage new device-technology support quickly

Company mergers often bring about network changes. You may be called upon to quickly implement a new network technology. The ideal scenario is for your vendor to become your partner. As partners, both of you should work together to identify the specifications necessary to manage a new technology—and implement that technology in days or weeks, not months or years. Look for vendors that offer technologies that allow you to build modules. It is easier and more effective to plug a new technology into an architected infrastructure instead of having to recode changes to an existing architecture.

9. OS and client should be independent

More companies than ever are reallocating stranded assets. This includes putting heavier loads on existing systems. Make sure your NMS solutions are portable to different operating systems. At the very least, make sure your solution(s) can run equally on Solaris and Microsoft servers. Linux-functioning solutions are becoming more popular. Clients running applications that hook into your NMS solutions need to be versatile and network-flexible. Whether you’re connecting to the NMS via LAN, WAN, or dial-up, the NMS client needs to be simple to install (a no-CD installation is preferable) with forward-thinking dependencies (such as Java) that allow different dependencies to support different applications. For example, you don’t want to install one NMS client and have it break a different NMS client on the same operator’s workstation.

10. Automate everything you can

The biggest challenge facing IT departments today is automating NMS solutions for a dynamic environment. Many companies still see their internal IT department as a heavy cost center—so much so that they have been outsourcing their technology needs and focusing on their core competencies. Automation can assist in these environments. It is, therefore, crucial that your NMS solutions be able to track moves, adds, and changes without manual intervention. Eliminate tools that cannot automatically track changes in your network.

For example, moving a server within the network should require no intervention. Similarly, your NMS tools should be the trusted source from which other systems are fed because of their ability to automatically track network changes. Other minimum requirements include the automation of tasks including auto-filtering events (fault), auto-tracking circuits (performance), and auto-update version numbers (inventory). The more convoluted your multi-vendor solution is, the more integration is needed and the less inherent is the automation process.


Today, no business stands still for long. Moves, changes, mergers, acquisitions, new markets, and new initiatives all play a role in matching IT investment with business priorities. Companies that understand how their network functions in a changing environment recognize the importance of change management and are more likely to extract value and productivity from their networks and IT departments.

Editor's note: This story was originally posted 9/30/2003 at Enterprise Strategies.