Monitoring alerts using the AWS SDKs or API - MediaLive
Tips for ListAlerts and ListMultiplexAlertsTips for ListClusterAlerts

Monitoring alerts using the AWS SDKs or API

You can use your preferred AWS SDK or API to monitor alerts about channel activity, multiplex activity, and activity on the nodes in an AWS Elemental MediaLive Anywhere cluster.

There is an operation for each resource:

  • ListAlerts for a specific channel, with optional state (SET or CLEARED) filter. The response includes information about the alert, the alert status, and the applicable pipeline.

  • ListMultiplexAlerts for a specific multiplex, with optional state (SET or CLEARED) filter. The response includes information about the alert, the alert status, and the applicable pipeline.

  • ListClusterAlerts, for a specific cluster, with an optional state (SET or CLEARED) filter. The response includes information about the alert, the alert status, and the node and channel that the alert applies to.

You can query for alerts from when the alert is first set until the alert has been in a CLEARED state for approximately 5 minutes.

Tips for ListAlerts and ListMultiplexAlerts

The information in the response for channels and multiplexes is the following:

  • Alert type: A short description, usually 3 to 4 words.

  • Messages: A long description.

  • ID: The ID is created when the alert is first set. Note the following:

    • The ID is a hash of a monotonically increasing count.

    • Each channel (or multiplex) has its own ID series that increments independently of other series.

      This means that two channels might each have an alert with the same ID during the same time period, especially if the channels start at close to the same time. Therefore in your code, you should assign a unique identity to each alert by combining the channel ID and alert ID.

    • The rollover of IDs for one channel (or multiplex) is reset each time the channel (or multiplex) restarts. Keep this in mind when performing a query.

Tips for ListClusterAlerts

The information in the response for clusters is the following:

  • Alert type: CLUSTER_NODE_HEALTH for every type of alert.

  • Message: A long description, which might include variables that resolve differently in each cluster or in each instance of an alert.

  • ID: A long description that is often equivalent to the message.