elig@bezeq.com wrote:
> You might need to adopt a "reactive" approach.
> The routers will send Netview interface-down traps if you configure your
> SNMP statements properly. You can then run an automatic script based on
> these traps, to store the information and process it later to make it
> presentable in WEB format. This way, you save all the extra polling traffic
> volume.
>
If you are relaying on agent traps, you have to remember that they are UDP
based, and are not guaranteed to be deliverd. Also, I tend to see a lot of
duplicates. NetView version 6 has SNMP based polling available. If you
aren't at Version 6 yet, you could, as a ruleset action to an interface up
trap, get SNMP information from the router interface that tell you the SNMP
System Uptime when the interface changes state. Subtracting this value from
the system uptime of the router will tell you how long ago the interface
changed state. If it is less than the polling interval, it was a probably a
real event.
--
Ray Schafer | schafer@tkg.com
The Kernel Group | Distributed Systems Management
http://www.tkg.com
|