What is Distributed Networking?

What is meant by distributed network management?

Today's local LAN architectures have transitioned from ****-enjoyment networks to switched networks to meet the ever-increasing bandwidth demands. The switched fabric has dramatically changed the way networks are managed. Traditional RMON probes viewed the entire network as a ****hedge network. To overcome this shortcoming, industry-leading network vendors have built RMON software into hubs, switches, and even network cards to collect data from associated devices and segments and transmit that data to a computerized network management system called distributed RMON or dRMON. RMON2 enables network management to be expanded to include application-layer protocols that provide first-time, real-time, business-class capacity planning and resource allocation capabilities.

Distributed network management includes a range of features and functionality, including the following:

Adaptive Policy-Based Management

Adaptive Policy-Based Management has the ability to set policies or rules by which the network responds to changes in network performance and security. It can store user and application profiles to determine their corresponding quality of service levels and bandwidth requirements. Adaptive policy-based management reduces the complexity of computer network management by giving end devices the intelligence to ask for what they need from network devices.

Distributed lookup and monitoring

Distributed lookup and monitoring refers to moving network management tasks, such as device lookup, topology monitoring, and status polling, from a network management workstation to one or more remote workstations. This will reduce the workload of the central management workstation and the amount of traffic traveling through the network backbone and WAN links.

With distributed management, workstations with network management software can be configured as either collection workstations or management workstations. Each collection workstation is responsible for collecting information from a specific set of users (called a domain) of a managed object. A management workstation can make network management functionality available to users through that workstation or indirectly through one or more management consoles. The network management console becomes a virtual access point for computer network management information as defined by a particular MIS operator.

Intelligent Filtering

In large-scale network environments, in order to limit the information load at a high level, distributed management uses intelligent filtering to reduce the amount of data. Intelligent filtering refers to selecting the level of information for further processing. Distributed management uses four types of filters to remove unnecessary data at different points of the system, namely Discovery Filter, Topology Filter, Map Filter, and Alert and Event Filter.

Distributed Threshold Monitoring

Threshold event monitoring enables computer network administrators to detect and isolate faults within the Domain where the problem occurs before they are felt by users. The ability to forward threshold events means that management workstations no longer need to collect events for all objects that require threshold monitoring. Collection workstations can independently collect SNMP and RMON trend data from their associated objects and activate Action-On-Event thresholds based on the data. Through a feature called Status Aggregation, the collection workstation makes the information available to other collection and management workstations that need it, and optionally forwards the data to a central console, allowing data collection levels to be used for forecasting, capacity planning, trending, and archiving of service level contracts, among other things.

Distributed threshold monitoring reduces the volume of services on the network and the burden on the CPU and management resources of the central management platform. Currently, some advanced vendors have SNMP-based software built into network devices that take on the distributed threshold monitoring tasks of the central network management software. Coupled with RMON tools included within the computer network management software, this built-in software collects diagnostic and performance data and performs specific validation operations independently of the central management console. This distributed RMON feature extends the monitoring and collection capabilities of the central network management system across switched LAN segments.

Dynamic Polling and Judgment Logic

Polling engines typically collect data at regular intervals. With dynamic polling and judgment logic, the polling engine is able to automatically and autonomously adjust the polling intervals to obtain more detailed information about the performance and operation of devices and network segments in the event of anomalies or network failures.

Distributed management task engines

Task engines are time-consuming, resource-consuming, and bandwidth-intensive transactions, and delegating task engines makes computer network management more automatic and independent. Typical distributed task engines include distributed software upgrades and configuration, distributed data analysis, and IP address management. Distributed management models with these features have the advantages of good scalability, low complexity, faster response time and better performance and accessibility of information.