With Cisco® Active Network Abstraction (ANA) software, Cisco Systems® delivers a radical new approach for managing service provider networks. Cisco ANA creates a virtualized, service-oriented network model that reflects the near-real-time state of every network element and maintains an end-to-end view of the entire network regardless of vendor or technology. It meets service provider demands for service-level management systems that truly reduce complexity and costs.
The oppressive cost of managing today's complex service provider networks is the direct result of poor integration between operations support systems (OSSs) and hundreds of individual element management systems (EMSs). With no shared information pool among these management systems, service providers are forced to perform redundant information-gathering processes and manually exchange information among applications. Service providers facing competition forcing them to introduce and assure complex new services find this process expensive, error prone, and time consuming.
This issue becomes immediately critical as service providers consolidate multiple, single-service networks into a single, multiservice network based on the Cisco IP Next-Generation Network (NGN) architecture. Provisioning services and monitoring the network to assure consistent service-level agreements (SLAs) and optimal uptime take on an unprecedented level of complexity as service providers struggle to combine multiple data, voice, and video services into a single management infrastructure. Traditional IP network management systems remain domain- and device-centric, and cannot cost-effectively scale to address the complexities of the Cisco IP NGN architecture.
Cisco ANA software represents the Cisco vision of an entirely new management architecture that enables end-to-end, service-level management in very large multivendor, multitechnology, multiservice networks. This long-awaited, elegant solution manages converged, multiservice infrastructures based on the Cisco IP NGN architecture. The solution is based upon a virtualized network model that creates service-level views, enabling rapid integration of existing provisioning, fault management, and billing systems. This customizable, integrated management platform vastly simplifies service provisioning, configuration, monitoring, and troubleshooting processes to reduce operational expenditures, shorten time to revenue, and increase customer satisfaction.
Challenges
Service providers sell services, but manage networks. Managing very large, multivendor IP networks is complicated and expensive. The current generation of management tools is not service-oriented, but domain- and device-oriented. Service providers have provisioning and troubleshooting systems and tools that do not scale indefinitely, and most service provider IT teams already suffer from their limitations. Implementing a service order (such as installing a video-on-demand service) using the appropriate combination of EMSs and site visits to provision and activate it requires deployment times of weeks or months. Troubleshooting complex multivendor networks is equally arduous, because management teams lack tools that offer end-to-end service visibility and root-cause analysis across multiple vendor products and domains. Faults can generate hundreds of alerts, but correlating and analyzing them to identify the root-cause failure (and fix it) can take hours or days.
Today's service provider networks are both massive in size and complex in nature. Recent mergers between behemoth telecommunication providers combine already huge networks, exponentially increasing their scale and complexity. This trend adds considerable urgency to addressing the challenges of managing these networks. A Tier-1 service provider network uses many elements from many vendors, which means an EMS portfolio can include hundreds of EMSs, all speaking different protocols and formats. Network elements from dozens of vendors each have their own EMS. There is no agreed-upon format for communicating between a particular element and the rest of the management framework; different elements from the same vendor can use different protocols and information formats. There is also considerable variation (and inconsistency) in application security and APIs.
Service providers want to automate their flow-through systems to gain efficiencies, protect their margins, and price services competitively. They need tools with end-to-end, service-level views that allow them manage services, not just network domains or devices. While many management systems include topology views, none of them present a comprehensive service topology that shows everything in the network. Expensively engineered systems-integration projects construct complicated flow-through systems by cobbling together applications that were not written with interoperability in mind beyond a variety of APIs. Networking vendors have made little effort to standardize northbound interfaces to simplify systems integration. Today's management systems may have topology capabilities, but they are limited to specific vendor equipment, technologies, network layers, or service instrumentation. Adding a single new platform to a service provider network today requires re-engineering of existing flow-through systems and middleware, making it slow and expensive to perform any kind of network upgrade or enhancement.
Another challenge is poor information sharing among management systems and applications. There is no central repository of network inventory and real-time state information that every system can access. Each system polls the network individually using distinct formats and protocols. The lack of integration between applications makes it difficult to identify faults and their impact on network and service performance. This situation wastes time performing multiple discoveries and can often lead to discrepancies between multiple sets of information. Without universal information sharing and end-to-end service topologies, operators have difficulty understanding how services are affected by new configurations, misconfigurations, new components, or component failures. Capacity-planning, provisioning, performance, and fault-management processes drawing upon inconsistent sets of information may create unpredictable problems. Both vendors and IT teams duplicate basic software functions such as installation routines, configuration engines, and persistent databases for reporting. Consolidating these basic functions into a single system used by all applications can save significant time and expense.
While service providers need fast time to market, rapid service provisioning, and service-level troubleshooting and maintenance, these goals remain unattainable until vendors deliver a customized, integrated management system that allows all management components to interact with a consistent interface to all network elements and management applications.
Requirements to Attain Service-Level Management
Managing services, instead of individual links, devices, and domains, allows service providers to closely align the services they sell with the operations that enable them. Examples of these services are the following:
• Enterprise VPNs based on Multiprotocol Label Switching (MPLS) and Metro Ethernet
• Managed customer premises equipment (CPE) services for enterprises and small to medium-sized businesses
• Consumer "triple-play" services that deliver telephone, television or video on demand, and Internet access to the home over a broadband service
• Consumer broadcast television over IP
Service providers want granular, end-to-end visibility and control in their IP NGN similar to what they had in their time-division multiplexing (TDM)-based voice networks. Yet visibility and control in a TDM environment was expensive to build and monitor. The Cisco IP NGN offers new revenue potential and efficiencies that deliver clear advantages over traditional TDM networks. Achieving those efficiencies depends on the industry's ability to deliver a comparable level of granular visibility and control through a fully integrated, optimized network management system. This new system honors and incorporates existing OSS/BSS and management systems, which represent multimillion-dollar investments. It simplifies the processes of adding new platforms or upgrading existing ones to improve business agility and network availability. It speeds troubleshooting with automated, accurate root-cause analysis.
Cisco Active Network Abstraction
Cisco delivers a radical new paradigm for managing very large service provider networks. Cisco Active Network Abstraction (ANA) enables a unified management system that delivers true, end-to-end, service-level management of service provider-sized networks. The most commonly used functionalities in the Cisco ANA software are the following:
• Real-time discovery and reporting of complete device inventory (physical and logical) and multilayer network connectivity
• End-to-end service tracing and monitoring (through network simulation of packets and traffic flows)
• Topology-based alarm correlation, root-cause analysis, and service-impact analysis
• Performance reporting and thresholding
• Cross-network service activation
• Real-time device configuration
Enabling complete integration of existing systems and new capabilities, Cisco ANA creates a common forum that communicates with all southbound EMSs to the network and all northbound applications, OSSs, and BSSs. It is a virtualized management layer that resides between the network elements and OSS management applications, mediating communications among them (Figure 1).
Figure 1. Cisco ANA Logical Architecture
The Cisco ANA layer normalizes information gathered from network elements and presents it in a consistent format to northbound applications. Management applications can access this information using an API type that best fits their functionality. By abstracting this communication, service providers gain the freedom to manage the network without having to upgrade management code every time they add or upgrade a device. It is now possible to add network elements from any vendor with relative ease, just by incorporating the EMS into the southbound interfaces of Cisco ANA.
But Cisco ANA is not merely an abstraction or application-mediation layer. It is based on a powerful, virtualized network model that reflects the near-real-time state of every network element and an end-to-end view of the entire network, representing all services, domains, and devices regardless of vendor or technology. This virtualized network is constructed to represent the four building blocks of a network: interoperable interfaces (physical and logical ports), protocol stacks, forwarding entities (switching and routing tables), and adjacencies (physical and service links).
The foundational components of Cisco ANA are Virtualized Network Elements (VNEs), which act as autonomous "mini-systems." Each VNE corresponds to a specific network device, discovering and holding information only about that device. It uses device-specific protocols to learn its physical and logical inventory characteristics, its configuration, its immediate peers, and its connectivity references. Every VNE knows its device's security access schemes, traffic engineering tunnels and trunk layouts, and service-based connections. It actively communicates with the device and maintains real-time state information. Collectively, VNEs contain complete and accurate inventory and connectivity information for the entire network.
VNEs viewed together comprise a virtualized network model that behaves much like the network itself. Unlike other system-level management systems that adopt an IP-level, single-box approach to network management, the virtualized network uses a parallel-processing, distributed-intelligence architecture, where each component contains localized information, just as its corresponding element does in the real network. Each VNE only performs local calculations, just as its counterpart does in the real network. Cisco ANA is a network approach to network management. Unique in its design, this architecture scales far beyond single-box approaches of other management systems, presenting an entirely new paradigm for attaining end-to-end, service-level management. It can accommodate tremendous growth to manage global networks with millions of subscribers. It represents the leadership, innovation, and quality that service providers should expect from Cisco-solutions that address business problems and open new revenue opportunities.
Adjacent to the VNE layer in Cisco ANA is the Cisco ANA Gateway, which allows Cisco ANA Workflow Clients and northbound applications to access the virtualized network. Cisco ANA Gateway presents a common portal through which network operators can perform element and link discovery, fault management, service activation, and performance monitoring and management, using applications available with Cisco ANA software. They can also interface with existing OSS and BSS applications to access their functionality. This Common Information Model builds service assurance with highly accurate, real-time state information that is available to all management applications through the virtualized network (Figure 2). Managing the real network through this virtualized network model enables tighter integration of subscriber-generated services for greater automation and control. Integrated business and subscriber information makes it easier to understand the impact of network and device failures on business and subscriber services.
Figure 2. Virtualized Network Topology Creates a Common Information Model
The Cisco ANA topology eliminates the need for complex business rules based on fuzzy logic and endless lists to identify root causes. It simplifies alarm correlation for instant root-cause analysis and service-impact analysis. Operators are no longer subjected to sorting through hundreds of alarms when a link fails. The topology model calculates the root cause. The operator can find it in a few clicks of the map and initiate the repair procedure. The operator can then determine which services and customers are affected by the outage and verify that the network is rerouting traffic around the failure.
The initial version of the Cisco ANA product family includes the following applications:
• Cisco Configurable Device Management Platform (CDMP)-Delivers multidevice, multitechnology, multivendor element management with unified monitoring and configuration
• Cisco Network Discovery and Cisco Service Discovery-Provide accurate, real-time discovery of network inventory, topology, and services, with service-level topology views that show physical, topological, and logical relationships
• Cisco Network Fault Isolation and Cisco Network Service Isolation-Provide extensive fault-analysis capabilities for rapid, accurate fault detection, isolation, and correlation at the network and service levels with accurate root-cause analysis
• Cisco Service Activation-Facilitates automatic configuration of relevant devices to activate a service regardless of vendor, model, or software version; performs activation and configuration in parallel with no single point of congestion
• Command and Threshold Builder-Defines and executes customized configurations for service activation and device-specific configurations
• Workflow Client-The interface through which operators access and operate Cisco ANA and its applications
Benefits of Service-Level Management
Consolidating all your management systems and applications into a unified management system based on Cisco ANA software yields the following benefits:
• Accelerate time to revenue-Cisco ANA enables end-to-end service activation in hours, from a single console. Customers can order the service in the morning and it will be available that afternoon. This capability increases billable service time and delivers a considerable competitive edge.
• Achieve global scalability-Using a network approach to create a network management system is vastly more scalable than single-box solutions. Each VNE only has knowledge of its peers. The virtualized network scales through the addition of VNEs, one for each network element. There is no risk of memory overflow because of the distributed, parallel-processing architecture of the model.
• Reduce MTTR, increase service availability-Service assurance is a faster, more reliable process through the intelligent fault correlation and rapid, automated root-cause analysis inherent to Cisco ANA; less time is spent diagnosing and isolating a problem, allowing technicians to fix them sooner.
• Increase service agility-End-to-end visibility of the entire network (all vendors, technologies, and domains) and a central information repository shared by all management applications support rapid upgrades, new platforms, and changes to services without costly systems integration, giving IT the means to respond quickly to business demands.
• Reduce complexity-The Cisco ANA layer unifies all EMSs, OSS, and BSS applications, enabling an elegant, simple approach to service-level management of multivendor, multitechnology networks through a single platform.
• Preserve existing investments-Cisco ANA allows service providers to continue using existing management applications and proprietary OSSs and be ready for rapid integration of future applications and device managers.
• Reduce capital expenditures-Requiring less integration and fewer management components.
• Reduce operational expenses-Cisco ANA speeds upgrades and incorporation of new platforms, accelerates problem identification and resolution, and allows fewer administrators to manage more services.
Cisco ANA Meets the Challenges
The Cisco ANA management platform offers a unique way to simplify service management over a multiservice network. This section discusses some common services and the role of Cisco ANA in service activation, monitoring, and troubleshooting.
Managed CPE Services
Service provider networks deliver managed CPE services using a variety of CPE from many vendors, across a variety of local connectivity technologies such as T1 lines, Frame Relay, Metro Ethernet, and DSL. The complexity of managing millions of connections today absorbs massive amounts of time, and it is impossible to bring profitability to these services with the amount of manual intervention required for each subscriber.
The Cisco ANA virtualized network vastly simplifies the technical issues and related costs of deploying managed CPE services to multiple customers over a single network. It specifically addresses the following issues:
• Scale, complexity, and diversity-Because the virtualized network is the database, the Cisco ANA approach scales directly alongside the real network, allowing operators to view and manage the complexities of multiple services and millions of customers in a multitechnology, multivendor network.
• Rapid network element adaptation-The abstraction capabilities of Cisco ANA eliminate the complexities of developing and integrating network adapters for every new network element (such as CPE), allowing providers to bring new customers, endpoints, and services on board rapidly and inexpensively.
• Overlapping IP address spaces-Cisco ANA eliminates this severe stumbling block for legacy service and revenue assurance systems with Network Address Translation (NAT) capabilities that allow operators to distinguish between customers.
• Event correlation and service-impact analysis-The sheer scale of today's service provider networks cripples the efficacy of traditional event-correlation tools. Cisco ANA has unique Implicit Network Imprinting technology that automatically correlates thousands of alarms and allows administrators to trace root causes and their impact on services within minutes.
• Integrating the OSS and BSS with the network-Synthesized network intelligence and the Cisco ANA distributed knowledge base enhance interaction between OSS and BSS applications to help guarantee service performance and dynamic, service-aware capabilities.
VPN Services with MPLS and Metro Ethernet
Service providers are migrating VPN services away from pure ATM networks to gain the economies of scale enabled through MPLS technology. Metro Ethernet is gaining popularity as a low-cost, high-bandwidth last-mile service. However, managing an MPLS core and multiple Metro Ethernet subnets using domain managers is costly and time-consuming. Cisco ANA adds critical, end-to-end visibility and control to the VPN service. Operators can view pathways on a per-subscriber basis from endpoint to endpoint, incorporating the multiple access and transport domains into a single, service-level topology that is easy to interpret and evaluate. For example, a retail bank may acquire a regional bank with 75 branches in a single metropolitan area. Cisco ANA allows the service provider to rapidly integrate the new branches into the existing network, usually without requiring any change in CPE or connection type. The service provider can add the new endpoints to the existing customer profile, and provision VPN services to all the new branches in days instead of months. The advanced fault management, event correlation, and service-impact analysis capabilities allow the service provider to attain and maintain a competitive SLA.
IP Broadcast Television
Customer expectations for high-quality broadcast television services are high. They are used to DVD-quality picture and sound, and are gaining familiarity with high-definition television. Traditional telecommunications companies seeking to enter the broadcast television market by delivering services over a Cisco IP NGN must meet strict quality standards or face failure. Standard, best-effort IP transport is insufficient. Latency, dropped packets, jitter, and temporary outages result in subscriber dissatisfaction and service cancellation. End-to-end visibility and highly granular control are essential for maintaining consistent, high-quality service delivery. With unequalled granularity of visibility and control, Cisco ANA delivers the robust network management needed to support broadcast television services. Operators can monitor and manage the service from a single subscriber-path viewpoint to regional or network-wide views to deliver the high level of service assurance required for customer satisfaction and retention.
Why Cisco?
Cisco Systems is the worldwide leader in networking for the Internet. It continues to earn its leadership position with its IP NGN vision for delivering multiple services over a single, intelligent network infrastructure. The keystone for successfully realizing the IP NGN vision is a scalable, fully integrated management system for the entire service provider network, not just the Cisco parts. The innovative paradigm of managing the network through a virtualized network model enables the real-time, service-level visibility and controls that service providers have wanted for years, but were not possible until now. The Cisco ANA products are part of a dedicated investment in a next-generation management system for the Cisco IP NGN architecture. Cisco is actively engaged with third-party developers and other networking vendors to assure the most comprehensive multivendor, multitechnology solution possible. As the leader in this cooperative effort, Cisco invites your comments and suggestions.
