Guardian 0.6.3 - RabbitMQ Alarms

The Guardian 0.6.3 release introduces functionality for AmazonMQ RabbitMQ brokers, nodes and queues.

Current defaults for brokers are:

• ConnectionCount Warn and Crit
• MessageCount Warn and Crit

Current defaults for queues are:

• MessageCount Warn

Current defaults for nodes are:

• SystemCpuUtilization Warn and Crit
• RabbitMQMemUsed Warn and Crit
• RabbitMQDiskFreeLimit Warn and Crit

Note that the default values for these alarms are based off the smallest instance type, mq.t3.micro. Adjust these values accordingly. For more information regarding AmazonMQ metrics: AWS Docs

Id corresponds to the broker name. Currently there is no MQ api that exposes the rabbit node names or queue names. These can be found by going into the cloudwatch console and looking at Node Metrics by Broker or Queue Metrics by Virtual Host. Alternatively, we plan on adding this retrieval functionality into monitorable

Resources:
  AmazonMQRabbitMQBroker:
  - Id: dev-rabbitmq
  AmazonMQRabbitMQQueue:
  - Id: dev-rabbitmq
    Queue: my-special-queue-name
    Vhost: /
  AmazonMQRabbitMQNode:
  - Id: dev-rabbitmq
    Node: rabbit@localhost

There have also been additional bug fixes.

• Fix when specifying ExtendedStatistic
• Fix redshift health status alarm setting off false alarms while healthy

Guardian 0.6.3