Systems

By Greg Berlocher

Since their inception, network management systems have made the metamorphic change from basic monitoring aids to feature-rich management tools. Built on powerful and robust computing platforms, today’s network management systems now can deliver a host of valuable services to corporations, carriers and governments, and the term management system no longer is a misnomer.

Network management systems (NMS) have been around for decades but their functionality was limited, to say the least. In reality, they should not have been called management systems but rather monitor and control systems (M&C). The original M&C systems emulated the front panel display of a piece of remote equipment on a computer screen. This representation provided an operator, or satellite engineer, a virtual look over the shoulder of field personnel to help piece together information regarding service anomalies or service interruptions. M&C vendors had to interface with a myriad of different hardware — both old and new — as users wanted to integrate every separate brand and model of hardware they had installed into the system. M&C vendors spent many years building up impressive libraries of device drivers which allowed their proprietary computer systems to talk to a disparate group of satellite hardware.

Convergence

Throughout the last decade, there has been a major escalation in computing power and improvements in transmission protocols and data exchange formats between computer systems. These developments have served as the building blocks for the newest generation of network management systems. Much has been written about convergence. "While satellite is extremely important, it is no longer independent from other technologies," says Mark Krikorian, COO for ILC, a manufacturer of network management systems founded in 1993. "We see ourselves now as managing converged, heterogeneous networks."

While this shift provides a wide range of benefits to end users, convergence has created a number of challenges to those who manage large, diverse networks. The blending of satellite and information technology (IT) groups, traditionally separate entities, is just one result. "The IT and media/video groups are now having to collaborate and learn more about the systems each other use," says Randy Palubiak, president of Enliten Management Group, a consultancy based in Marietta, Ga. "More and more, our clients utilize both satellite and high-speed data circuits to deliver content. Different groups now need to have some familiarization about what the other group does."

Roger Franklin, COO of Crystal Computing Corp., agrees that convergence on network operations is having a significant impact on M&C systems. "We see a tremendous interaction now between the IT and video distribution groups," Franklin says. "IT may be monitoring MPEG video on the corporate [wide area network] but wants to keep an eye on the satellite network as well."

Enhanced Features And Capabilities

Comparing the features and functionality of today’s powerful network management systems to earlier M&C systems is like comparing a broadband VSAT terminal to a 12-foot backyard TV receive-only antenna. The advanced capabilities allow corporate and government users to actually manage converged networks using tools that until now were lacking.

"Network management systems seem to atrophy over the years," Krikorian says. "Many clients tell us that their existing NMS was the right choice when they bought it but now it is outdated." ILC’s philosophy is to go beyond monitoring and control into the complete automation of networks through a high-level architecture, Krikorian says. "We believe that you must create a common model for network devices to communicate," he says. "Regardless whether they are simple or complex, the reporting model for all devices must be consistent. By using the same reporting model, we can then utilize any inputs from dry contacts to SNMP (Simple Network Management Protocol) to XML (Extensible Markup Language) to object oriented interface protocols." ILC’s Maxview operates under Oracle, Microsoft’s SQL Server, or the open source database MySQL. The system will run on Linux, Windows or Solaris operating systems. "We are a management chameleon," Krikorian says.

Contrasting the management capabilities of Maxview compared to earlier M&C systems, Krikorian points to ILC’s automated discovery and automated topology rendering capabilities. Automated discovery seeks out devices attached to the network and catalogs them, freeing humans from documenting — and sometimes missing — network components. Automated topology rendering simplifies, or eliminates, the need for humans to draw individual graphical representations of network devices. "A large client may have several thousand different network management screens and traditionally each screen had to be created individually," Krikorian says. "Maxview can now create these screens for you. The labor cost to create thousands of screens is a very definite cost of ownership which can’t be overlooked." ILC also has graphical driver development kit which allows clients to write their own drivers, Krikorian says. "We have a library of more than 3,000 drivers but clients always seem to need something new. The development kit uses a wizard-like interface so you don’t have to be a programmer to develop your own code," he says.

The ability to correlate alarms and events with specific circuits and services flowing across a network is another advanced feature earlier M&C systems lacked. "If a hardware component fails, the services running across it will fail also," Krikorian says. "Correlation allows network operators to quickly identify all the services that were affected by a hardware problem and quickly restore them. … The ability to correlate problems in the video world is especially important. The quality of a video signal may suffer without a network alarm being triggered. Signal jitter, blockyness, freeze frame and noise are all problems that creep up. The ability to monitor when the video quality of service drops below an acceptable level and then quickly identify the video path and associated equipment through cross correlation is very powerful," he says.

Crystal Computer Corp. has been building management systems for 20 years and always has placed a heavy emphasis on control. "Users need to be able to respond quickly to what they see on the management screen," Roger Franklin, Crystal’s COO, says. "Human control has always been important but automated control of network components is a reality now that computers have become so powerful. That allows automated fault recovery, no matter where the equipment is located."

The company’s Crystalvision NMC employs a distributed software architecture. "You need centralized control of a diverse network but centralized control doesn’t have to be from a single point," Franklin says. "Our architecture is a peer-to-peer network of servers. This may be dozens, or even hundreds of servers in a large network. There is a substantial amount of logic required to hand off control of a device from one server to another. … "After Hurricane Katrina, business continuity became a significant issue with network operators. It always bubbles up to the top in the scope of importance with clients. We used to discuss disaster recovery as a side topic, but now every client wants to discuss business continuity and the real need for redundant centralized control," he says.

Crystalvision NMC is based on proprietary software, which means "our system runs faster and is much more secure from viruses," Franklin says. "We don’t use a central database because it would slow the whole system down, plus there are synchronization issues that would have to be addressed. The whole process of switching control centers with Crystalvision NMC takes only a few seconds."

In keeping with Crystal’s long-term goal of complete network automation, the company was in final preparation to release their new predictive failure analysis tool at press time. Crystalvision NMC will now be able to sample inputs from network equipment, store them in a database for trending, look for threshold violations, make predictions on anticipated failures and generate trouble tickets based on the expected severity of the failure. "In the past, M&C systems only told you when something failed, making you completely reactive," Franklin says. "Now network operators can rely on their management system to tell them they are about to have a problem. Further, it will let the network operations team know if the problem needs to be addressed immediately or if it can wait a few days. Managers can be far more effective in their utilization of personnel with this tool."

Advanced M&C systems also can present the data in different ways to different users, Franklin says. "There are many different groups in an organization that are affected when a network has problems," he says. "Not all of them care about the same things. We can present physical views of equipment for network engineers, rack views for maintenance personnel, and service views for network operators. This helps everyone focus on what is important."

Integrated carrier monitoring is another powerful feature that earlier M&C systems lacked. Crystalvision NMC has the capability to do spectrum recording and playback. The service operates in the background and works in conjunction with a spectrum analyzer to check for anomalies. "We can then synchronize the data monitoring and carrier views to see if an RF glitch caused a particular video problem."

Scheduall, is used by more than 1,200 media clients and 400 broadcasters in 39 different countries. The company has solutions that help clients with resource management, planning, personnel utilization and cost control. Roger Kleckner, vice president of strategic business development for Scheduall, says that computing standards used today helps their company compliment network management systems made by others. "Scheduall’s products are more synergistic with NMS systems than direct competition," he says. "Since new NMS systems can interface with Scheduall’s systems via XML, we can now have a bidirectional dialog with those products," Kleckner says. "This opens up a whole new world of relations between different vendors."

"Our preferred approach is not to take absolute control of the network but to interface with a network management system, through an existing software interface. The heavy lifting has already been done. We prefer to stay complimentary and continue to add value. If a network management system is smart enough to accept external commands, we can interface with it."

Scheduall has added a number of features that early users of M&C systems did not offer. Cost control, scheduling and utilization capabilities now are a reality. Web- deployable applications further disperse critical information to the people who need to know. Clients can now book airtime online; freelancers can check their schedules; and significant alarms and events can be sent via text message or e-mailed to Blackberries and other hand-held devices.

Kleckner noted that end users are driving the development of many new features found in new generation network management systems. "They are asking: How come I can’t have this?" he says. "Our clients know where the bottlenecks are in their operations and have a good idea what capabilities we have in the software to help. The new features we have developed are purely a collaborative effort."

Conclusion

Convergence has changed the way organizations transmit video and data and has added to the complexity of managing networks. It is not hard to imagine a time when every major network will use multiple technologies for managing and delivering content to end users. It is clear that organizations will need advanced network management capabilities that go far beyond what has been available to date. Fortunately, several vendors have had their finger on the pulse of the market for quite a while and are delivering powerful systems that can meet the needs of the future.

Greg Berlocher is a twenty-year veteran of the satellite industry. He is a writer and photographer and his work has appeared in numerous publications.