Uploaded image for project: 'OpenNMS'
  1. OpenNMS
  2. NMS-7918

Slow BridgeBridgeTopologie discovery with enlinkd.

    XMLWordPrintable

Details

    • Bug
    • Status: Closed (View Workflow)
    • Blocker
    • Resolution: Fixed
    • 16.0.3, 16.0.4, 17.0.0
    • 18.0.0, Meridian-2016.1.0
    • Enhanced Linkd / Linkd
    • Security Level: Default (Default Security Scheme)
    • None
    • OpenNMS runs an three dedicated server, one for the Database, one for the GUI and one for the monitoring. The server have more than enough memory and CPU power so this couldn't be the problem.

    Description

      The BridgeBridgeTopologie discovery in a big installation of an customer ( over 7000 nodes and 12000 links) is very slow and need sometimes hours for a single node.
      Picture 1 shows the amount of time that enlinkd needs for the BridgeBridgeTopologie the have 34 active layer2 interfaces. The other Protocols (cdp and lldp) are as fast as usual.
      In a test run with only one Node enlinkd needs only seconds for the discovery (Picture 2).
      I add the logfiles of enlinkd, eventd and the output.log.
      The JDBC-Errors in eventd.log seems to be caused by reimport form a big amount of Nodes with the REST-API.

      Attachments

        1. 16.0.4.tar.gz
          3 kB
        2. 2.PNG
          2.PNG
          17 kB
        3. enlinkd.log
          40.36 MB
        4. eventd.1.log
          100.01 MB
        5. eventd.2.log
          100.00 MB
        6. eventd.log
          29.69 MB
        7. features_17.0.0-enlinkd.tar.gz
          3 kB
        8. output.log
          5.04 MB

        Activity

          People

            rssntn67 Antonio Russo
            AndreasFuchs Andreas Fuchs (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved: