A Final Word

It is increasingly difficult for enterprises to countenance throwing out all of their legacy hardware and management software in order to install the latest device offerings. There is a need for allowing such users to migrate slowly and steadily towards the packet-based networks of the future. The cost of hardware is falling, and for this reason good-quality NMS provide a degree of vendor product differentiation.

Another interesting aspect of consolidation and ongoing procurement is that SP equipment inventory will become increasingly similar. Service providers will find it more difficult to differentiate their services using hardware alone. NMS will offer a useful means of gaining competitive advantage. These considerations apply to both network operators and NE vendors .

The central role of MIBs in network management has been a major theme of this book, and we hope that MIBs now hold no surprises for readers. Vendor and standards organizations can do much to promote manageability by creating well-designed MIBs. Wherever possible, standard MIBs should be used.

Pushing more intelligence into NEs can be readily accommodated with MIBs such as the FTN MIB discussed in this chapter. It is possible that specialized networking hardware such as the network processors from Intel and IBM may be required for such MIBs. However, the pattern is clear: NEs will become increasingly sophisticated and capable of autonomously executing highly complex management functions. This will help to improve the scalability of networks that encompass many (i.e., hundreds or thousands) such devices.

The running example used in this book was MPLS. This was done to provide an interesting backdrop for the NMS discussions and also for a number of other reasons:

• MPLS allows for a connection-oriented layer 3 network.

• Phased migration of layer 2 technologies (such as ATM) to layer 3 becomes feasible .

• Layer 2 skills can be de- emphasized .

• Emerging standards such as PWE3 and Ethernet over MPLS pave the way for generic cores, moving complexity to the network edge.

• Consolidation of multiple technologies may also help in reducing the number of incompatible management systems.

Even if MPLS is not deployed in a large enterprise network, the benefits of NMS are considerable:

• An overall network perspective is provided by the NMS.

• NMS provides centralized management rather than using numerous EMS and proprietary products.

• It becomes possible to proactively manage the network using policies, that is, damage prevention rather than damage control.

The need for solutions in network management technology is a challenge for software developers. It becomes necessary to acquire a working knowledge of many different technologies, including IP, MPLS, ATM, Frame Relay, and SONET/SDH. The linked overview technique described in Chapter 3 may be of assistance in rapidly getting a handle on these different areas. While there is rarely a substitute for experience, a willingness on the part of software developers to learn new technologies quickly can help vendor companies in shoring up skills shortages. This should have a direct and positive impact on product revenues .

On the commercial side, the global economic downturn (that started around March 2000) has forced most enterprise/SP network operators to assess their options. Investment does not tend to occur unless there is a proven return to be made.

As vendor organizations return to the path of profitability, it will become essential for them to produce good-quality differentiated solutions. This will translate into products that generate the cash needed to provide training for crossfunctional cooperation. It is only the latter that will permit the long- term solution development and maintenance capability needed for the emerging NEs. Customers may be able to assist in this process by providing some of their own engineering capability, thereby extending crossfunctional cooperation outside the vendor organization. The creation of value is the difference between the cost of producing and selling solutions.

In Chapter 7, we saw how straightforward it is to rapidly create NMS building blocks. The available development tools (e.g., Visual C++ and JDMK) are very easy to use and are supplied with reasonably good sample code. These provide base-level components that can be incorporated into larger bodies of NMS function ”that is, FCAPS. The really hard problems are, as usual, related to scale, though usability and generic software are also crucial.

While vendors and enterprises have their problems, we also should not forget SP operator problems. The TeleManagement Forum [TeleMgmtForum] has reported that adding a new NE to an SP network can cost in excess of $20 million. This is most likely due to some combination of:

• NMS changes required for the new hardware and associated NMS modules

• Interoperability problems with existing devices

• Firmware bugs in the new devices

• Integrating management for the NEs into existing OSS workflows and business practices

Similar costs apply to large enterprise networks. Many technologies, such as MPLS, are implemented long before the standards are complete. This is a necessary part of vendors keeping up with their competitors . Competitive advantage may go to those vendors whose products match the adopted standards. SNMP is an established standard that is widely deployed. Developers of NMS and NEs can use standard tools such as UML and SDL in conjunction with standard programming languages to create increasingly open systems. By open , we mean that UML and SDL allow for the development process to be opened up to all stakeholders. This can result in a better mapping between user requirements and deployed solutions.

Security is critical to successful network management, and SNMPv3 facilitates this. The SNMPv3 security model allows for extensions if necessary. It is likely that 3DES, AES (Advanced Encryption Standard), and their eventual successors will be required.

Network management is a very broad, exciting field. The trend toward favoring solutions over technology puts network management in a prominent position. The industry equation is simple: Good-quality NMS technology will help network operators to provide reliable, high-performance networks that meet organizational needs.