Skip to:

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

Issues

Select view

Select search mode

 
16 of 16

enlinkd logging hibernate errors (lack of unique index)

Fixed

Description

I think I started seeing this after upgrading to v32.0.0, but that might be a coincidence.

In this case, the duplicate rows were identical except for a fraction of a section difference in isisnodecreattime - I manually removed the duplicates and I don’t see the error anymore.

The isiselement table should have a unique constraint on nodeid. From opennms mattermost chat, this also affects the lldpelement table.

Acceptance / Success Criteria

None

Attachments

2

Details

Assignee

Reporter

Labels

HB Grooming Date

HB Backlog Status

Components

Sprint

Fix versions

Affects versions

Priority

PagerDuty

Created October 20, 2023 at 2:00 PM
Updated January 31, 2024 at 3:04 PM
Resolved January 25, 2024 at 9:06 AM

Activity

Show:

Christian PapeJanuary 25, 2024 at 9:06 AM

Merged.

Christian PapeJanuary 24, 2024 at 7:36 AM

Please review:

  • PR:

Christian PapeJanuary 23, 2024 at 11:51 AM
Edited

There are two different problems in this issue:

  1. the nodeId column in the tables LldpElement, CdpElement, IsisElement and OspfElement needs to be unique. I created the liquibase changeset for that.

  2. it seems that the NullPointerException for the given topology is caused by a broken SNMP implementation. I looked through the code and the class OspfIfTableTracker is querying for several OIDs to determine the local interfaces. This somehow doesn't seem to work for the given topology:

    So, only one entry is returned and the OID is not increased to return the next one. So, Enhanced Linkd encounters for a device a link to another device but cannot determine its local interface for this link. I added null-checks to avoid the exceptions, but this is something that should not happen in a normal environment.

Ronny TrommerJanuary 16, 2024 at 4:01 PM

I have started to troubleshoot the problem within a lab environment. Enabling just LLDP in Enlinkd was not a problem and the topology was discovered correctly. I have just enabled OSPF discovery and got an unknown exception on the node page. The error message in jetty log is attached in the ospf-exception.log.