Network Monitoring and Management System

To ensure consistent network performance, you need to constantly monitor and manage any network faults and configurations. As the size of the network increases, it becomes difficult to monitor resources and configuration. This in turn increases network vulnerabilities such as performance slowdown and failures. To monitor the performance of network resources when the size of the network increases, you need network management tools. To troubleshoot network problems, you need to perform:

Fault detection: Refers to the initial tasks to be performed when a network starts witnessing problems.

Configuration management: Refers to the set of tasks performed after diagnosing the problem area in the network and re-establishing the configuration of network components during and after problem resolution.

Performance management: Refers to the set of monitoring tasks for monitoring the prospective areas of network problems and taking remedial actions.

Resource management: Refers to the tasks that provide Quality of Service (QoS) to network users and applications. You need to prioritize the applications or end users that require more resources as compared to other end users or applications.

The currently available network management solutions provide detailed drill-down reports about the performance of network components. You can use these reports to detect the reason of network failure or performance slowdown.

Network monitoring is an effective tool to determine network capacity, establish a standard for network performance, and generate alerts when network problems occur. A performance standard is established by taking into consideration parameters such as network performance and traffic patterns on the network over a period. These statistics help derive an average rate of network performance. You can use this performance standard to compare the current network performance and check for any deterioration. Some network monitoring tools that offer remote and local network monitoring services are: CiscoWorks, RMON, and MIBs.

Some of the Cisco IOS commands used for collecting network-related information are debug diagnostic commands and show process cpu commands. The debug command runs in the privileged exec mode on the router console. It provides information about various router operations, such as traffic generated or received by the router on the network and errors on the router interface. The only limitation of this command is that it consumes a significant amount of network resources for the router. This leads to performance slowdowns for the routers. Hence, it is recommended that this tool be used only occasionally to determine specific network problems. The debug command can be used for a specific facility, task, or protocol that could be the cause of the problem. For example, if a particular protocol in the TCP/IP suite requires debugging, the debug command provides the facility to use the event or packet option. Event debugging consumes fewer resources as compared to packet debugging. A sample output of the debug command is displayed in Figure 4.1.

Figure 4.1: The output of the debug diagnostic command.

In addition to the debug command, there is a show process cpu command, which determines the amount of CPU utilization. This helps you identify router processes that are using the CPU inefficiently and, as a result, enable optimal CPU utilization. Figure 4.2 displays the output of the show process cpu command.

Figure 4.2: The output of the show process cpu diagnostic command.

Configuration management tasks are usually performed manually. If you have made changes in the network configuration, ensure that the network components are concurrent with the latest configuration setup. For example, if a printer is removed from a network or if the IP address of the printer changes, update the static IP address of the new network printer on each system to ensure printer access over the network.

An Application layer protocol that facilitates exchange of management information between network devices is SNMP, which is a component of the TCP/IP suite. SNMP allows you to manage fault detection and recovery of your network. Its components include the SNMP agent and the SNMP management station. The SNMP agent is deployed on the router, and the SNMP manager queries this agent to collect network statistics. This way, you can determine how the network is functioning and anticipate any problems that may occur in the future. The SNMP framework is displayed in Figure 4.3.

Figure 4.3: The SNMP framework has two managed devices.

As shown in Figure 4.3, SNMP commands perform different tasks:

Get Request: Gets information about the attributes of a managed object from the SNMP agent

Get Next: Performs the same task as the get request command for the next object in the managed device tree

Set: Sets the value for a particular managed object attribute

TRAP: Alerts the SNMP manager about events on the managed object

Get Bulk Request: Obtains a large volume of information through a single operation

Inform Request: Enables the SNMP manager to communicate with other SNMP managers

Response: Retrieves responses from the agents for the Get Request, GetNextRequest, GetBulkRequest, SetRequest, or InformRequest Protocol Data Unit (PDU)

There are four major constituents of the SNMP framework:

Managed devices: Are present within an embedded SNMP agent

Management server: Connects with managed devices, providing various services to management applications

Management protocol: Uses the SNMP protocol for receiving and sending encoded messages for the SNMP manager and agents

Management information model: Defines the managed resources within the SNMP framework using a pseudo object-oriented design in which all objects are stored virtually in MIB

An example of a scalable network management product developed by Cisco for service provider networks is the Cisco WAN Manager (CWM). It performs all the major functions of a network management system (NMS), including performance, configuration, and fault management. It provides statistical data for network management and stores it in the database. In addition, CWM easily integrates with existing network management and operating systems. While monitoring the network, you can use the CWM interface and tools to configure devices and track network statistics. As a result, you can monitor and manage the entire network from a central location.

Besides CWM, other products that can be used for performing partial network monitoring functions include CiscoView and Cisco DFM. CiscoView is a device management tool that provides dynamic status and configuration for Cisco devices. Cisco DFM proactively detects network faults and alerts network managers.