Versions Compared

Old Version 2

changes.mady.by.user Sara Weisz

Saved on

compared with

New Version Current

changes.mady.by.user Sara Weisz

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

Configuring Analytics Offload to DPOD

  • API-C Connect v10

    • Login to the analytics VM:

      Code Blockssh <user>@<ANALYTICS_SERVER_IP_ADDRESS> sudo su -

      Open an SSH session to a server that can execute kubectl commands against the Analytics subsystem.

    • Edit the AnalyticsCluster CR and replace add the following content under spec -> external -> offload with the following content:

      Code Block
      kubectl edit analyticscluster

spec:
  external:
    offload:
      enabled: true
      output: |
        http {
          url => "http://<DPOD_IP_ADDRESS>:60020/apicanalytics"
          http_method => "post"
          codec => "json"
          content_type => "application/json"
          id => "offload_http"
        }

  • API-C 2018Connect v2018

    • Log in to the API-C Connect Cloud Manager.

    • Go Navigate to Configure Topology → Analytics Service → Click on the 3-dots icon and choose Edit → Advanced Analytics Configuration → Ingestion.

    • Edit the Outputs.yml and , enable the offload and save it.

      Code Block
      offload_output_enabled: true

    • Edit the Offload_output.conf with the wanted offload type and save it. and add the following content:

      Code Block
      output {
  if "apicapievent" in [tags] {
    http {
          url => "http://<DPOD_IP_ADDRESS>:60020/apicanalytics"
          http_method => "post"
          codec => "json"
          content_type => "application/json"
          id => "offload_http"
        }
    }
}

    • Save the changes on the Edit Analytics Service page.

    • Login to the analytics VM:

      Code Block
      ssh <user>@<ANALYTICS_SERVER_IP_ADDRESS>
sudo su -

Troubleshooting

  • Offloading is processed by the *-ingestion-* pod pods.

  • Delete the pod to make sure it picks up the new configuration (sometimes it fails Sometimes the ingestion of pods fail to reload the configuration if it is they are still retrying to send old messages to an old endpoint). In such a case, delete the pods to make sure they pick up the new configuration:

    Code Block
    kubectl get pods --all-namespaces
kubectl delete pod *-ingestion-<ID>
    [Optional] To troubleshoot, look at the pod 
    *

  • Use the following command to view the pods logs:

    Code Block
    kubectl logs *-ingestion-<ID>*

Associating/unassociating DataPower API Gateways to Analytics Service

  • Log in to the API-Connect Cloud Manager.

...

  • Navigate to Configure

...

  • Topology → Click on the 3-dots icon of one of the DataPower API Gateways → Select “Associate” or “Unassociate”

...

  • .

...