Chapter 18. Network Management Basics

Network Management Basics > FCAPS Model Chapter 18. Network Management Basics This chapter covers the following topics: FCAPS Model Network Management Architecture Network Management Protocols An Introduction to TMN Network management describes the methodology used to manage and maintain network operations and respond to user requirement changes. With the implementation of Simple Network Management Protocol (SNMP), local area network (LAN) and wide area network (WAN) components can be monitored and managed, often from what is called a central Network Operations Center (NOC) facility. The network management platform is an integrated suite of functions that can be implemented on one machine, or it can be implemented on several machines or databases spanning thousands of miles, supporting several organizations. Some examples of a network management platform are HP Openview or CiscoWorks 2000. Network management can be segmented into two categories: Tactical Tactical network management relates to proactive and reactive situations, such as network failures, congestion, and unacceptable service quality. The tasks include troubleshooting, configuration, and adjusting traffic flows. Strategic Strategic network management involves a long term perspective that is oriented toward adequate planning to avoid shortages as the network grows. Strategic tasks use information to adjust operations, optimize quality, and manage facilities to reduce overall operational costs. This chapter discusses network management models that are both tactical and strategic. This includes a discussion of the following: FCAPS model Network management architecture Network management protocols Telecommunications Management Network (TMN) FCAPS Model The ISO FCAPS model is a major contributor to network management. It is similar to the OSI Reference Model used for internetworking (for more information regarding the OSI model, see Chapter 2, "OSI Reference Model"). The FCAPS model is made up of the following components: Fault Management Detects, logs, notifies users of, and (if possible) automatically fixes network issues. Most fault management systems poll the managed objects for error conditions and present this information to the network manager. Fault management identifies and isolates network issues, proposes problem resolution, and subsequently logs the issues and associated resolutions. The testing of any fault resolutions

should be performed on non production systems prior to deployment in a "live" network. Configuration Management Monitors network and system configuration information so that the impact on network operations (hardware and software elements) can be tracked and managed. Network changes, additions, and deletions need to be coordinated with the network management personnel, often in a NOC. Accounting Management Measures network utilization parameters so that individual or group users on a network can be regulated, billed, or charged. Performance Management Measures and makes network performance data available so that performance can be maintained at acceptable thresholds. Security Management Controls access to network resources as established by organizational security guidelines. Most network management systems address security regarding network hardware, such as someone logging into a router. Where FCAPS is a model for network management, the next section discusses the architecture for network management. Network Management Basics > Network Management Architecture Network Management Architecture The architecture of a network management platform is made up of a common set of relationships and structure that exists between managed devices and a management entity, as illustrated in Figure 18 1. Figure 18 1. Network Management Topology The following list details the components found in a network: End stations Computer systems and network devices with software that enables the sending of alerts (to management entities) when network issues are identified; for example,

any LAN server or user's workstation configured with network management software. Management entities These entities are programmed to react to end station alerts by executing a predefined set of actions, such as the following: Event logging System shutdown Console notification Automatic attempts to repair the system (such as reload or reboot) The elements of a management system are as follows: Network manager agents Integrate with network devices, platforms, and applications Network manager applications Network management servers that collect network management data from the network manager agents and correlate the data for analysis by the network administrator/manager Network management agents carry network management data and report network transmission problems to a network manager. The network manager controls a set of management agents and ensures that these agents collect the appropriate information. Network management agents are software modules that compile information about the devices within which they reside and automatically or manually respond to all polls. Agent information is conveyed to the management entity within the Network Management System (NMS) via a network management protocol, such as Simple Network Management Protocol (SNMP). Network Management Basics > Network Management Protocols Network Management Protocols Network management protocols carry network management data between the managed devices and the management console. The following network management protocols are defined by various Internet Engineering Task Force (IETF) standards: SNMP The SNMP is used to communicate with a management "agent" in a network device. A remote manager collects status information and controls the remote device through the SNMP agent. There are three versions of SNMP: SNMPv1 Reports only whether a device is functioning properly. In other words, it tells whether the device is "up" and running or "down." SNMPv2 SNMPv1 with the addition of security and Remote Monitoring (RMON) Management Information Base (MIB) support. The RMON MIB provides continuous feedback to the network manager without having to be queried by the SNMP console. SNMPv3 SNMPv2 with the addition of message level security. SNMPv3 also includes an MIB for remotely monitoring and managing the configuration parameters for the SNMPv3 Security Model. SNMP MIBs The SNMP agent delivers the MIB information to the network manager. Each managed resource has an MIB, containing what can be known about the device and the device's capabilities. An MIB for a router contains information about each interface

bandwidth speed, protocols supported, and current status. An MIB for a server has information about the Central Processing Unit (CPU), operating system, memory, and disk usage space. MIB2 A standard MIB defining basic interface information such as bandwidth speed, numbers of packets sent and received by the interface, numbers of broadcast and unicast packets, and errors. RMON An MIB that controls an agent monitoring a single LAN segment. Collects information as instructed by the network management console about traffic levels, which systems are talking, and specific conversations between two parties. RMON2 An MIB for controlling agents that monitor traffic across the network (LAN/WAN). RMON2 measures traffic flows between different parts of the network and identifies which protocols and applications are in use by each system. Common Management Interface Protocol (CMIP) OSI standard protocol used with the Common Management Information Services (CMIS); CMIS defines a system of network management information services. CMIP provides improved security and better reporting of unusual network conditions. CMIP was proposed as a replacement for the SNMP but has not been adopted by the networking community for widespread implementation because of SNMP's incumbency. Network Management Basics > An Introduction to TMN An Introduction to TMN TMN is the International Organization for Standardization (ISO) and International Telecommunications Union (ITU) standard detailed in the ITU T Recommendation Series M.3000. The M.3000 series defines the framework for the planning, provisioning, installation, maintenance, operations, and administration of telecommunications networks and services. This framework is flexible, scalable, reliable, inexpensive to run, and easy to enhance. TMN provides for more capable and efficient networks by defining standard ways of doing network management tasks and communicating across networks. TMN also allows processing to be distributed to appropriate levels for scalability, optimum performance, and communication efficiency. NOTE The full ITU T M.3000 recommendation series can be found at wwwcomm.itsi.disa.mil/tmn/tmn_itu.html. The principles of TMN are incorporated into a telecommunications network, sending and receiving information and managing network resources. Telecommunications networks are made up of switching systems, circuits, terminals, etc. In TMN terminology, these resources are referred to as network elements (NEs). TMN enables communication between operations support systems (OSS) and NEs.

TMN architecture and interfaces build on existing open systems interconnection (OSI) standards, which include but not limited to the following: CMIP Defines management services exchanged between peer entities (other network management devices). Guideline for Definition of Managed Objects (GDMO) Provides templates for classifying and describing managed resources. Abstract Syntax Notation One (ASN.1) Provides syntax rules for data types, such as those found in an MIB. OSI Model Defines the seven layer OSI Reference Model. NOTE The TMN standards work has been incorporated into the work by other standards bodies, such as the Network Management Forum (NMF), Bellcore, and the European Telecommunications Standards Institute (ETSI). The Synchronous Optical Network (SONET), SONET Interoperability Forum (SIF), and the Asynchronous Transfer Mode Forum (ATMF) are specifying TMN compliant management interfaces. Summary The ISO FCAPS model is a major contributor to network management. It is similar to the OSI Reference Model used for internetworking. FCAPS is an acronym for Fault Management, Configuration Management, Accounting Management, Performance Management, and Security Management. TMN is based on the OSI management framework. Management functions are performed by operations comprised of CMIS primitives. Processes that manage the information are called management entities. A management entity can take on one of two possible roles: manager or agent. Network manager and agent processes send and receive information (requests and notifications) using CMIP. The architecture of a network management platform is made up of a common set of management entities end stations and manager stations (used for the analysis of collected network management data). Frequently Asked Questions (FAQ) 1:Which is better, FCAPS, or TMN? A1: TMN has become separated into two parts: TMN as architecture for telecommunications management and TMN as a set of supporting protocols. Initial TMN proponents and implementations did not differentiate between architecture and protocol and often ran into adversity. The open standards concept behind TMN was widely accepted, although the

underlying protocols were not readily adopted. The TMN model helps to categorize, prioritize, and specify the responsibilities of telecommunications management products and services. For the first years of TMN's existence, the industry was focused on standard protocol implementation, particularly at the element and network management layers. The future of TMN, however, is not in the protocol domain. TMN will continue as a framework for telecommunications management, but it will become less focused on defining protocols to support the TMN architecture. FCAPS is the methodology used to implement the TMN standards for network management. Whereas TMN defines how networks, such as ATM, can be managed, FCAPS is the model used to implement the network management architecture. There is no clear answer regarding which is better, FCAPS, or TMN. As the industry has evolved, TMN has found a niche within the FCAPS model. 2:Is it necessary to monitor and manage every network device? A2: No, it is not necessary to monitor and manage every device on a network. Device monitoring and management depends upon the mission critical nature of that device. For example, it can be considered more worthwhile to an organization to monitor and manage a WAN router or corporate Internet/intranet servers rather than to monitor/manage individual workstations. Although it is desirable to have visibility into and manage as much of the network as possible, even at the desktop level of granularity, there is a balance between the size and scale of the network and the network management organization itself. It is considered unreasonable for a single network manager to manage a large ATM WAN to the desktop level of granularity (such as routers, switches, servers, hubs, and LAN workstations), when management of the routers and servers is a huge task in and of itself. Case Study This case study illustrates how a network might be managed. Figure 18 2 depicts a simple threenode Frame Relay WAN. This Frame Relay WAN is managed by a third party (neither the customer nor the network service provider) network management entity.

Figure 18 2. Network Managed Frame Relay WAN This third party entity requires a Frame Relay Permanent Virtual Circuit (PVC) to each managed site so that it can have visibility into managed devices at those respective sites. The network manager uses SNMPv2 so that device status can be collected without having to be polled from the network management console, conserving network bandwidth by minimizing the amount of network management traffic and utilizing the manager console for data collection and analysis rather than as a polling server. This same network manager console provides the network manager remote access to the managed devices at each site for change management, such as configuration or router software (IOS) upgrades.