Previous Table of Contents Next


The NMS Middleware

In addition to selecting HPOV as the SNMP-based network management application, the organization selected the following third-party value-added applications to develop an enterprise level network management system:

  Remedy Corporation’s ARS, as the event-tracking software for generating, tracking, and documenting trouble tickets.
  A RDBMS from Sybase Inc., which works with the ARS to store and distribute event tickets and reports.
  A performance-analysis application from SAS Inc., which integrates with HPOV to access MIB data to maintain historical data, produce summaries, perform statistical analysis, and generate network reports.
  Isicad Corp’s Command 5000, as the physical and asset management software to obtain a graphical representation of the physical location of devices on the network to track network assets and device-to-device connectivity.
  Frontier’s NETScout RMON probes, to collect networking data on remote LANs and forward selected information to the network management system.
  A redundancy application from Qualix Group to monitor all of the network management applications running on the primary and secondary NOC workstations.

The organization carefully considered including additional types of value-added applications, but in maintaining a phased system implementation approach, instead it chose initially to roll out a suite consisting of the barest minimum needed to manage the network effectively.

The Enterprise Network Management System

Physically, the NMS comprises four SSPARC 20 WS that are situated at the NOC. The primary workstation and redundant secondary workstations (i.e., the WS1 and WS2) are connected to the TCP/IP network by an Ethernet LAN. These workstations collectively run HPOV NNM, ARS clients, and Sybase clients. Two additional workstations, WS3 and WS4, support the NMS. WS3 houses the SAS application as well as the ARS clients and Sybase servers. WS4 hosts Isicad’s physical management software. WS4 is also used as an X-Terminal.

The four NMS workstations are connected internally to each other by an Ethernet LAN, which is also used for printer connectivity. Each workstation on this LAN supports Ethernet and FDDI interfaces and are also connected to high-speed, secured modems through the RS232 port to perform out-of-band management. A standalone 486 PC hosts the capability to manage the modems that are used for the out-of-band network access.

The SNMP-Based Management Platform

The Hewlett Packard’s OV NNM 3.3.1 is used to monitor and control the network and provides the necessary platform to integrate middleware applications to maximize data sharing and processing. The NNM:

  Automatically discovers the devices on the TCP/IP networks and monitors the status of these devices.
  Automatically draws the topology of the network based on the discovered information and creates appropriate map views. A map is a graphical and hierarchical representation of the network.
  Collects performance information from the device’s MIB, stores it for trend analysis, and graphs the collected data.
  Defines event thresholds for MIB objects remotely.
  Takes specific actions upon receipt of specific SNMP traps.
  Diagnoses and displays network faults and performance problems.
  Allows integration with the popular third-party applications to enhance the system’s management capabilities.

HPOV IMPLEMENTATION

Two copies of NNM are deployed on NOC workstations WS1 and WS2. WS1 houses the primary NNM, whereas WS2 hosts the secondary copy. Each copy is configured and customized in an identical manner. Database synchronization between these copies is maintained by scripts that are automatically initiated on a periodic basis.

Each NNM is configured at various levels. The map views of each are customized to view and isolate faulty devices easily. Each NNM’s alarm and event handling is configured to notify operators in real-time using a combination of beeps, pop-up windows, and E-mail messages. The map and event color schemes are carefully defined to quickly determine fault criticality (e.g., red implies critical failure; yellow is minor; blue is unmanageable). The capability to obtain real-time performance graphs for selected backbone links is also implemented for quick status reports. Finally, to ensure that sufficient Management Information Base data is collected for performance, fault, and accounting purposes, selected MIB variables, with associated collection frequency, are identified for all network devices. Applicable threshold checks are also instilled.

SAS Reporting Software

The SAS Corporation offers a suite of products capable of integrating with HPOV NNM to perform data processing, analysis, and report generation. These products collectively allow the users to import data from NNM MIB files into its proprietary database, reduce the collected data, and perform statistical analysis on this data to generate strategic and performance baseline reports. These reports are used by the NOC operators, managers, and sustaining engineers to jointly perform network trend and capacity planning analysis.

By using the SAS product suite, the organization is able to parse information and reduce the amount of storage needed without compromising on disaster recovery requirements; also, SAS provides the ability to more easily summarize network performance data and generate reports on circuit utilization, LAN utilization, and device performance.

Trouble-Ticketing Software

The organization has developed a help-desk call process model with the goal of handling the user’s problem on the first call. To that end, the organization will people the help desk on a 24 * 7 basis with a mix of operators and operations support personnel. These NOC personnel interface with the remote node personnel using a combination of commercially available trouble-ticketing packages, secured telephone lines, facsimiles, and E-mail to resolve problems in a timely manner.

Remedy Corporation’s ARS 2.0 was selected as the centerpiece for the organization’s help desk. The ARS generates trouble tickets and tracks network events in a customizable work-flow process. The ARS uses an interactive process similar to E-mail to track network events with repair personnel. To accomplish this, the ARS provides the capability using a GUI to customize the following:

  Defining information to be passed among repair personnel.
  Defining automatic processes to assist operator data entry on event tickets.
  Notification and communication of defined information among repair personnel.
  Define work flow rules such as escalation procedures.
  Provide statistical analysis and reporting capabilities.


Previous Table of Contents Next

Copyright © CRC Press LLC