Skip to:

  1. Skip to Jira Navigation
  2. Skip to Side Navigation
  3. Skip to Main Content

The SnmpPoller is sending the interfaceOperDown events every time a requisition is synched or a single node is rescanned.

Description

This problem is really easy to reproduce:

1) Enable the SNMP Poller on service-configuration.xml and restart OpenNMS.
2) Add some nodes to a requisition
3) Add a policy to enable the SNMP Poller on all the SNMP interfaces (for facilitate the tests).
4) Be sure that the target node meet the conditions for sending the event for at least one interface (operStatus=down + adminStatus=up)
5) Sync the requisition

After synch the requisition, you should see an interfaceOperDown event for every interface that match the condition (operStatus=down + adminStatus=up).

Every time you sync the requisition or send a forceRescan on the node in question, you are going to see the interfaceOperDown event en each interface that match the condition.

In certain occasions, the event is sent twice (which is even more weird).

Acceptance / Success Criteria

None

Attachments

2
  • 28 Apr 2013, 07:41 PM
  • 26 Apr 2013, 12:25 PM

Lucidchart Diagrams

Activity

Show:

Alejandro Galue June 25, 2013 at 9:03 PM

Antonio has helped me to fix the problem.

Fixed on revision d9d0642039ec6aaaf3f8e37e2e49f28ec3c613b7 for 1.10

Note: If OpenNMS is restarted and the current status of the interface is down, the event is going to be sent no matter if it was sent before.

Alejandro Galue May 10, 2013 at 6:23 AM

If the EnhancedLinkd is a priority now, you can work on the SnmpPoller after that.

The thing is right now, on the stable 1.10.9, the SnmpPoller is unusable because of this problem, and there are customers that depends on this.

Antonio Russo May 10, 2013 at 5:28 AM

Alejandro I'm working on EnhancedLinkd with the objective of finishing very soon. Then I have to work on snmppoller.
But let me understand how urgent is this problem....I can give it a priority on the work I'm actually doing

Alejandro Galue May 10, 2013 at 3:48 AM

Antonio, please don't forget this 😉

David Hustace April 30, 2013 at 10:58 AM

I have reviewed this issue with Alejandro. We should do the following:

Provisiond/Caspd should update the values of the snmpoperstatus and snmpadminstatus fields only when the snmppoll field is set to 'N' and Provisiond should be fixed to NEVER send any status events. It is not a poller.

Fixed

Details

Assignee

Reporter

Components

Fix versions

Affects versions

Priority

PagerDuty

Created April 26, 2013 at 12:24 PM
Updated January 27, 2017 at 4:20 PM
Resolved June 25, 2013 at 9:03 PM

Flag notifications