A Study from Seagate Software on their NerveCenter Product
The subject of this study is an international consulting organization and their use of NerveCenter for internal network operations. The purpose of this review is to outline specific benefits gained by customers through the use of Seagate Software's NerveCenter.
Background
The customer is managing a national network with one central Network Operation Center (NOC) and five regional support centers. This internal network consists of approximately 250 servers supporting Lotus Notes and file servers, 275 hubs, 130 routers and 66 switches.
Defining The Problem
One of the customer's main issues prior to implementing NerveCenter was a need to replace an out-dated platform management solution. Support was ending rapidly and they had outgrown its usefulness.
A second major issue resulted from the customer's research into event management software. They found that other event management solutions reviewed, provided simple on/off status polling for devices which would result in numerous and erroneous error messages and create false pages when attempting to alert technicians to network problems. The customer desired a system that had the intelligence to sort through many different types of network problems and notify operations staff when truly critical problems occurred.
Benefits from Use of NerveCenter
The network management platform chosen was Hewlett Packard's OpenView. The customer also chose the new NerveCenter 3.0 for NT due to a need for automated downstream alarm suppression. Voluminous quantities of network alarms were being generated and the customer desired a system to sort through these many messages to determine truly critical issues.
In a further effort to improve network efficiency and reduce network costs, the customer determined they could quickly realize the benefits of distributing this management further out into their enterprise by utilizing the strengths of NerveCenter's native Windows NT, client/server architecture.
By utilizing the distributed features of NerveCenter and placing event management on local servers, the customer was able to reduce, and in some cases eliminate, network polling over their expensive wide area network. The polling of network devices over the WAN used expensive communications resources which could otherwise be made available to provide end-users faster network access.
Another benefit of distributed event management was achieved when router links were disabled. The local management of devices was allowed to continue even if a router link was disabled between the impaired device and the central network operations center.
The customer was able to integrate NerveCenter to trouble ticket software from Remedy. A key NerveCenter 3.0 feature is NerveCenter to NerveCenter communication.
NerveCenter was configured to receive events from other NerveCenter implementations through the new trusted "informNC" connection. NerveCenter was able to correct many network problems automatically, and dramatically reduce the number of automatic trouble tickets generated. This allowed Remedy to perform paging functions and maintain a central point of dispatch.
One of NerveCenter's integration points is its ability to launch trouble tickets, update the ticket once it's open, and close the tickets automatically based on network events and activities.
Remedy was then configured to perform paging of technicians to manage the resolution of the trouble ticket from a central point of dispatch, thereby reducing redundant management tasks and ticket tracking. This resulted in improvements to network problem resolution response times.
The customer has implemented a Lotus Notes email application; the BMC Patrol agent was selected to monitor the servers running the Lotus applications.
All network traps and alerts generated from the application agents are forwarded to NerveCenter. Trouble tickets are then created, updated and closed automatically by NerveCenter.
An on-going problem developed within the network supporting the Lotus application. The NT servers would periodically stop for no apparent reason. NerveCenter behaviour models were quickly built to not only react to the alerts coming in from the agent, but to also periodically check to ensure the agent itself was running. If the check failed, NerveCenter automatically restarted the agent.
NerveCenter functions as the front-end of the customer's network management solution. All events and polling are done from NerveCenter, which then sends only appropriate error conditions on to Remedy and OpenView. This customer has enjoyed such a high level of success with their event management implementation that the senior IT management has decided to deploy NerveCenter on an international basis for the entire corporation.
