Data Network & Network Management

• A data network is a collection of devices and circuits used to transfer data between computers or devices. This allows geographically diverse users to share resources, like ATMs.
• Network design intends to meet the communication needs of organizations while increasing productivity and staying within budget.
• Comprehensive planning is achieved through suitability, reliability, scalability, and durability.
• Network engineers are responsible for installing, maintaining, troubleshooting, optimizing, and expanding networks and resolving issues.
• Network Management helps engineers deal with data network complexity according to management principles.

Network Management Process

• A process controlling complex data networks to maximize efficiency and productivity.
• It uses the ISO framework for network management that includes:
  • Configuration
  • Security
  • Performance
  • Accounting
  • Fault management

Configuration Management

• A process to gather information about the current network environment.
• Information is used to modify configured network devices, storing data to maintain an up-to-date inventory of all network components and reports.
• The methods of data collection are manual collection and auto-discovery/auto-mapping using ICMP, ping, traceroute and Network Management Protocol.
• Once configuration management information has been obtained, it will usually need to be updated.
• Network devices contain modifiable parameters, like:
  • Routers (routing tables and network interfaces)
  • Servers (application services and operating systems)
• Methods of data storage:
  • Unstructured e.g. ASCII files
    • Advantages are that they are easily read, accessed remotely, and administered.
    • Disadvantages are that they are inefficient and unable to provide complex data relationships.
  • Structured e.g. DBMS
    • Advantages - Stores data efficiently
    • Enable various types of information to one another
    • Versioning
    • Disadvantages - Need query language to access data (e.g. SQL)
    • Administrative overheads
• Provides central storage location for its data.
• It can automatically gather and update data, configure devices remotely, and facilitate network inventory reporting.

Security Management

• Security involves protecting sensitive information on a data network by controlling access points.
• Consists of identifying, finding, securing and maintaining secured access points.
• Security examples include the 1988 internet worm, and more recently ICMP and TCP-based DoS attacks.
• Increases confidence in using the network, with maintained security offering alternative practicalities by setting security management.

Performance Management

• Performance ensures networks are accessible and congestion-free.
• Monitoring network devices and links determines utilization and error rates.
• Helps provide consistent QoS by ensuring capacity is not overtaxed and impacting performance.
• This is context-specific.
• Involves several steps:
  • Collecting data on device/link utilization.
  • Analyzing through statistical analysis and workload modeling.
  • Setting utilization thresholds.
  • Using simulation to maximize performance.
• Important performance terms include:
  • Availability
  • Bandwidth/Throughput
  • Propagation
  • Congestion
  • Latency
  • Threshold
  • Utilization
• Reduces congestion/inaccessibility to provide service to users and helps examine network trends.

Accounting Management

• A process gathering network statistics to help network engineers decide resource allocation.
• Also involves gathering user data, setting quotas, and billing users.
• Allows effective measurement and reporting of accounting information and increases understanding of user utilization.
• Helps make informed decisions about allocating resources to ensure fair network access.

Fault Management

• A process of locating problems or faults on a data network.
• Detecting symptoms that may lead to problems and isolating the cause of symptoms.
• There is correlation between potential problems.
• Alarms do not include exact fault locations.
• Corrects the problem if possible.
• Gathers information for fault management through:
  • Interrupt driven critical events, but solely relying on such events not facilitating effective fault management.
  • Polling, finding faults in a timely manner.
• Manages higher bandwidth consumption and polling can be implemented using ICMP messages (ping).
• Enhances network reliability by providing tools to facilitate rapid fault detection, isolation and recovery and maintains complete network connectivity.

Network Management System (NMS)

• NMS comprises an underlying architecture and platform, with software for managing many network devices and applications.
• Example NMSs include:
  • Commercial implementations like HP Openview, SunConnect SunNet Manager, and IBM Netview.
  • Freeware implementations like Net-SNMP and OpenNMS.
• The NMS platform includes:
  • A user interface
  • A network map
  • A Database Management System (DBMS) / Management Information Base (MIB)
  • A query language
  • A customizable menu system
  • An event log
• Includes Centralized, Distributed, and Hierarchical architectures can be used for functionality.