Is everybody happy?

Used to be, keeping customers satisfied wasn't so hard. If they had dial tone when they picked up the phone, they were generally happy. But with the proliferation of IP and optical networks and service level agreements (SLAs), keeping people happy these days is much more of a challenge. Customers expect more from their service providers and they are demanding SLAs that meet specific standards. They are calling for best-effort service to be replaced by consistent, quality service.

This external pressure is compounded by a network that has been growing increasingly complex. More complicated technologies, additional infrastructure, diverse offerings and multiple partner relationships make delivering services to the customer — and monitoring their delivery — much more difficult.

“Services were straightforward and simple, and it was rare for carriers to have service level management,” says Terry Lindsey, chief technology officer for Orillion. “Now, a single service includes multiple components with services that can act sequentially or in parallel and have multiple protocols passing through their stacks. Service providers have to keep track of where each protocol is managed and understand the various services so they can configure each port.”

New management tools are using diverse strategies to give carriers useful information about each network device and about how customers' traffic is traversing through those devices. Some combine inventory information with traffic data that can be used for capacity planning and network development. Others use simple network management protocol (SNMP) to poll devices for performance monitoring and SLA verification. All offer service providers useful information about their networks.

“Instead of finger-in-the-wind analysis, management tools give carriers a cleaner picture of their networks,” says Jeremy Tracey, president and co-founder of Entuity. “Service level management is an underserved market. Carriers have critical needs in this area due to the massive capital expenditures they have made to expand their networks.”

Used effectively, these tools can provide much needed insight into a provider's network as well as the information necessary to lessen customers' worries and keep everyone happy.

Diagnosing the optical network

Several companies are focusing on providing management tools for optical networks. Optical's complex nature and newness requires a different type of management tool that addresses unique characteristics.

Clear Communications was so influenced by optical's growth that it refocused its mission and fine-tuned its fault and performance management software. The Clearview suite of products takes static network information from the inventory database and marries it with dynamic information about the optical switching paths. Clearview gathers the information about the different flows and provides a picture of the traffic's path through each circuit (Figure 1).

Using this information, Clearview displays service configurations, identifies service problems and the affected customers and consolidates performance details into reports that can be measured against SLAs.

“Managing optical networks at the customer level is challenging,” says Tom Richards, chairman and CEO of Clear. “We let service providers proactively manage the network by identifying the network's core competencies at the customer level.”

To expand its management capabilities, Clear is integrating Digital Lightwave's remote optical test and monitoring product to diagnose and isolate dense wave division multiplexing (DWDM) problems. Digital Lightwave's monitoring agents, embedded in the network access devices, will enable Clear to access DWDM analysis information and include it in network reports.

“Digital Lightwave's agents allow us to look at the DWDM performance metrics. They provide the nitty-gritty fiber optic parameters that affect payload,” says Richards. The partners have loosely integrated the two products and expect to have the combined solutions available by mid-summer.

AI Metrix is offering optical management tools with NeuralStar, its operations support system (OSS) suite of products. With experience in broadband and DSL networks, AI Metrix sees optical network providers as another sector of the carrier population that will need an integrated OSS solution.

“Before IP and optical, service providers were able to add, create and manage services manually. But they are almost completely reliant on network management systems now,” says Pete Oliver, president and CEO of AI Metrix. “The multiple routing and layered fabrics that each service takes require management tools that track each circuit's traffic.”

For service level management, AI Metrix offers NetAdvisor, which provides a single point to view all events and messages that affect a customer's service. NetAdvisor polls edge devices, gathering data to compare with a customer's SLA. It compares the bits received by the customer to the SLA to ensure that the service provider is meeting the contract.

“Providers with large subscriber bases and multiple devices don't understand what is happening to each customer. Our tools poll the devices and allow service providers to see which customers are affected when a pipe is lost,” says Oliver.

Although not limited to optical providers, Quallaby is winning customers in the optical sector. Quallaby's Proviso allows providers to integrate billing, provisioning and fault management data with data collected from network devices (Figure 2).

Proviso pulls traffic metrics from network devices every five minutes. This data, stored in the data mart, can be separated to highlight each customer's usage. Inventory information can be combined with device and customer information to decide what services are available to customers and how the customer's traffic is affected at each device.

“We allow carriers to provide consistent service and aggregate information on who is using which services and where,” says Mike Combs, director of product marketing for Quallaby. “We alert service providers before there is an outage so they can move customers off certain network segments before an SLA is violated.”

GiantLoop Network, which provides enterprise optical networking services, will use Proviso to collect data and monitor its physical and logical connections. The provider will also use the data to provide performance information and service reports to its customers.

Tiered services for IP networks

Allot Communications is concentrating its efforts on IP networks, specifically virtual private network (VPN) services. In its initial deployments, wireless, satellite and cable providers are using Allot to limit consumers' usage. In the next stage, Allot expects providers to use the tools to manage platinum, gold and silver services (Figure 3).

Sitting at the network edge, Allot supports up to 4000 customers. The software gathers how many bits are passed to the customer, notes the types of applications being used and marks the traffic with DiffServ values.

“Service providers can use our metrics for usage billing or to measure if SLAs were delivered. We provide information about latency, jitter, how many bits were transferred and more,” says Azi Ronen, executive vice president of technology and marketing for Allot.

ISP Warp Drive, a division of US Cable, is using Allot's NetEnforcer to allocate bandwidth usage to its customers. Rather than manually configuring each router based on the committed information rate to limit customers' usage, systems administrator Chris Pallack can manage subscribers though the Allot box.

“Allot allows us to manage all our bandwidth in one place,” says Pallack. “We can specify a customer's usage at the network level rather than at the hardware level, which streamlines the service management process.”

ViewGate Networks is also targeting service providers offering IP VPNs, but its software tracks frame relay and ATM traffic, too. Inteligo uses SNMP to poll the network devices for specific data, rather than bulk statistics, to decrease the overhead. The data is then correlated into performance reports for the service provider and the customer.

“Fault management is good at finding the root cause, but it can't show who was affected,” says Charlie Gallucci, vice president of marketing for ViewGate. “In the past, customers could not be sure if they were receiving the service that they paid for. We provide performance-monitoring reports to the service provider and customer that give customers confidence in virtual services such as VPNs.”

Cable & Wireless has been using ViewGate since October 1999 to provide performance reports to about 50 managed services customers. ViewGate polls ATM and frame relay switches providing reports on network usage, discarded traffic and throughput.

“Inteligo lets customers know what type of services we are delivering, and it helps them with capacity planning, says Neil Wilkinson, product development manager for data and IP reporting at C&W. “It gives us a real-time view of availability, and it alerts us if we exceed our thresholds and can't meet our SLAs.”

Opticom's management system also collects asset information, capacity and performance data. While other management tools poll network devices and gather information using SNMP, iView collects data from management systems such as Hewlett-Packard's OpenView, Aprisma Management Technology's Spectrum and Cisco's NetFlow.

Opticom recognizes that relying on other management systems for this data limits it capabilities. “Our universe of partners needs to evolve so we can control our own destiny. We are partnering with companies such as RiverSoft, so our customers won't need multiple management systems,” says John Morency, senior vice president of marketing for Opticom.

iView gathers four data sets from the underlying management system. The asset management data includes information about each switch, router, ports, serial number and so on. Usage information reveals SNMP traps, alerts, and outages on subnets. Performance data includes information on throughput and transaction response time. Last, it collects circuit capacity.

Entuity provides another management tool that can track fault information, performance data and inventory. Eye of the Storm provides Layer 2 visibility, performance monitoring, and reporting and inventory control to the port level.

“We provide real-time fault information that shifts network management from reactive to proactive,” says Tracey. “We show where the problems are occurring in the infrastructure and how to fix the problem before it impacts end users.” Eye of the Storm's Layer 2 monitoring and real-time fault isolation can be used to create more detailed SLAs and ensure that carriers' deliverables meet the contract, Tracey adds.

Broadening management tools' role

Retail providers complain loudly to wholesale carriers about the blind spots that affect their customers' traffic. Congested traffic in the last mile — or on another providers' backbone — defeat SLAs and lead to finger-pointing from the service providers that must honor the customer's contract.

To stop the finger-pointing, service level management developers would like to see carriers open their networks — or portions of their networks — to their partners. The same tools used to monitor relationships between enterprise customers and carriers could be applied to retail and wholesale relationships.

Carriers have resisted sharing network performance data because it would mean giving out customer and network operation information to potential competitors. That resistance may be wearing down, though.

‘Service providers can use our metrics for usage billing or to measure if SLAs were delivered. We provide information about latency, jitter, how many bits were transferred and more.’
— Azi Ronen, Allot Communications

“During the next 12 months to 18 months, we will see more next generation providers and competitive local exchange carriers partnering to extend their footprints,” says ViewGate's Gallucci. “SLAs and performance monitoring tools will play a key role in building these relationships and maintaining customer bases.”

AI Metrix' Oliver agrees that shared network data will go a long way toward improving customer satisfaction. “Carriers can't transmit alarms on a business-to-business basis, which makes it difficult to manage services. They need alarm messages that can be used for customers and trading partners.”

ViewGate, AI Metrix and others in the sector expect that their tools will be used in these relationships and at peering points. “Service providers must monitor their peering points and have SLAs with their partners at these points. Performance monitoring tools will allow service providers to react quickly, moving traffic off networks during spikes, outages and degradation,” says Quallaby's Combs.

Selling these monitoring capabilities is another potential use — and revenue stream for carriers — for the management tools. The reports and diagnostic information could be sold downstream to partners who could, in turn, sell it to their customers.

Management developers expect reselling diagnostic information and business-to-business partnerships will be two areas where carriers will adopt their products. But before the tools are accepted, the developers know service providers will have to accept new roles and forge new partnerships.

“Carriers closely guard what is going on in their networks. They are geographically limited and often sell virtual services that span multiple carriers,” says Gallucci. “Our products can monitor end-to-end connections. But the carriers would have to agree to open up their networks.”

An alternative to opening the entire network may be exposing only portions of the network. “Carriers are interested in communicating aspects of their traffic streams without handing over the keys to their castle,” says Combs. “They will need more mature SLAs, enabling technologies and new business relationships to make true end-to-end service management available.”

And with the result of the recent economic boom being massive buildouts, service providers indeed need the management tools that can help them live up to their SLAs and retain customers. And it's up to service providers and vendors to work together to find a way to provide not only the best service but the best SLA from end to end.