Re: Large Scale Netview Implementation Design

1999-01-12 14:55:14
Subject: Re: Large Scale Netview Implementation Design
From: Art DeBuigny <[email protected]>
To: nv-l AT lists.tivoli DOT com
Date: Tue, 12 Jan 1999 13:55:14 -0600

Let me comment on your plan, and tell you what I would do.

First of all, Maps can be transported between servers (regional and HQ), but
you have to take the entire thing.  Regional maps would still have to have the
full fledged  database if they were to be placed on the HQ box (with access to

I suppose it would be possible to have a seperate database repository for each
region, but it'll take some time to load each database and restart the daemons
and GUI on your central HQ server.

Is all your monitoring being done from the HQ, or only the critical?  If its
dopne regionally, with critical at HQ, then I'd install a client off of each
regional server, and have them running in your command center.  If its all done
centrally, I would replace the regional servers with Mid Level Managers.  Your
regional servers can be smaller, and less powerfull, and its much easier to
maintain one Server and 9 MLM's than it is to maintain 10 NetView Servers.  I
manage 3 Servers and its not fun.  :)

Art DeBuigny
debuigny AT dallas DOT net

Zahir Tohir wrote:

> I am required to design a large scale Netview implementation :
> The requirement is :
> 1) Need 1 Central Netview Server - which have all the maps of regional
> servers and all below it.
> 2) Need 9 Regional Server  - each regional maitain the view of its own
> Regional nodes only (map of its region)
> 3) Regional Netview server do polling on each region an update Central HQ -
> so that it minimize network traffic.
> 4) Regional Netview server filter all traps & event and forward only the
> CRITICAL event to the Central Netview.
> 5) All will be on AIX Machine
> First of all, can it be done ? I never done it before. Need help on this
> please...

<Prev in Thread] Current Thread [Next in Thread>