Juniper Apstra Cloud Services Edge Post-Setup: Config Changes and Troubleshooting

Configuration Changes

The Juniper Apstra Edge instance collects important information such as Username, Password, and IP Address of the Edge instance during the onboarding process in the Juniper Apstra Cloud Services UI. It stores this data in the cloud database for retrieval through Portal API (PAPI) endpoints. This is crucial for maintaining a stable connection and ensuring that the Edge container can communicate with the Edge instance.

Connection Failures

Connection failures are monitored by the Edge controller, which sends periodic status updates, including connection statuses, to the Juniper Apstra Cloud Services UI. These updates are important for identifying and addressing south-bound (Edge to instance) and north-bound (Edge to Apstra Cloud Services connection failures.

Juniper Apstra Edge Version Upgrade

The Edge container also plays a role in managing version upgrades by retrieving anomalies from the Edge instance upon requests from the alert-collector service. This means that the Edge version must be current to ensure accurate API requests for anomaly data collection.

Handling Direct Modifications

In scenarios where the Juniper Apstra Edge instance credentials are changed without updating them in the UI, the Juniper Apstra Edge container does the following:

• Detects disruption: Upon detecting a password change, the Edge container’s connection to the Edge instance is disrupted, prompting an immediate status update to the UI.

• Attempts to reconnect: The Edge container attempts to re-establish the connection up to three times before halting.

  In this instance, manual intervention is required. You must update the new password in the Juniper Apstra Cloud Services UI, triggering a config-push from the edge-collector service.

• Re-establishes connection: With the new credentials, the Edge container attempts to reconnect, continuously retrying until successful or a new config push occurs.

  A successful connection leads to a status update in the UI. A failed connection results in continuous retry attempts until the connection is re-established or a new config push is received.

Crash Recovery

If the Edge instance crashes, a recovery process occurs. Critical Edge information is stored during the startup event and should be backed up outside the host machine. This information is then used in the docker-compose.yml file for the Edge container during recovery, allowing for authentication with Juniper Apstra Cloud Services and the continuation of anomaly data forwarding from the Juniper Apstra Edge instance to the cloud.