Dear Adam,

 

Shortly after that, observium detected that device went down due to SNMP being unreachable. Which isn’t true…

 

Question is : considering that the device is basically doing nothing, why would the SNMP walk fail ?

 

Thank you for the feedback.

 

Best regards.

 

 

 

 

Youssef BENGELLOUN - ZAHR - Consultant Expert
Prodware France
T : +33 979 999 000 - F : +33 988 814 001 - ybzahr@prodware.fr
Web : prodware.fr

         

 

De : observium <observium-bounces@observium.org> au nom de Adam Armstrong <adama@memetic.org>
Répondre à : Observium Network Observation System <observium@observium.org>
Date : mercredi 30 novembre 2016 14:42
À : "observium@observium.org" <observium@observium.org>
Objet : Re: [Observium] Salvo of false alerts

 

This happens when a device fails to correctly respond to an SNMP walk. 

 

adam.

 

Sent from Mailbird

On 30/11/2016 13:37:12, Youssef BENGELLOUN - ZAHR <ybzahr@prodware.fr> wrote:

Dear Community,

 

I received a salvo of false alerts from an HP switch for all its’ interfaces :

 

 

 

Alert checker is pretty standard, it raises an alarm if an interface using over 90% of its’ bandwidth :

 

 

 

 

 

If I look at the eventlogs from the device, I can see a number of events (common to all interfaces) indicating the following :

 

2016-11-30 14:11:20   Interface changed: [ifAlias] X.X.X.X' -> NULL; [ifHighSpeed] '1000' -> NULL; [ifPromiscuousMode] 'true' -> NULL; [ifConnectorPresent] 'true' -> NULL

 

then, a few minutes later after the poller runs a new cycle :

 

2016-11-30 14:16:05   Interface changed: [ifAlias] '' -> 'X.X.X.X'; [ifHighSpeed] '' -> '1000'; [ifPromiscuousMode] '' -> 'true'; [ifConnectorPresent] '' -> 'true'

 

In the end, all alarms cleared out.

 

 

I’m curious to understand why observium would react in such a way. Any one could shed some light ?

 

Thank you.

 

 

 

Youssef BENGELLOUN - ZAHR - Consultant Expert
Prodware France
T : +33 979 999 000 - F : +33 988 814 001 - ybzahr@prodware.fr

Web : prodware.fr