Lightweight message queue and broker service that gives disparate services a common communication point.
RabbitMQ is an open-source message broker that supports multiple protocols. It acts as a broker, fielding messages sent from a “producer” application and distributing them to the “consumer” applications for which they are designated.
The New Relic RabbitMQ agent monitors the performance of RabbitMQ instances, giving insights into their activity. It targets the AMQP RabbitMQ distribution that integrates with Node.js.
RabbitMQ quickstart highlights
The New Relic RabbitMQ quickstart automatically instruments your RabbitMQ application, and comes pre-built dashboards visualizing:
- Total Used File Descriptors by Node
- Total Memory Usage by Node
- Consumers by Queue
- Consumer Message Utilization by Queue
- Total Messages by Queue
- Total Message Throughput by Queue
- Published Messages by Queue
- Published Messages Throughput by Queue
New Relic + RabbitMQ - Your tool for better monitoring
Monitor RabbitMQ to ensure it’s coordinating messages between producer and consumer apps in a timely and efficient manner. Silent failures are common in RabbitMQ, frequently causing messages to hang or never be delivered.
This can drastically warp the downstream behavior of an app. By monitoring RabbitMQ application health, system administrators can catch these errors before they occur.
The key metrics to keep an eye on include memory usage, message throughput and utilization by queue, and queue consumers. Monitoring memory usage allows for detection of system overloads. Similarly, throughput and utilization trade off on queue capacity.
Queues that are underutilized can be identified for message acceptance, and queues with low throughput can be investigated for hanging messages. Finally, insight into queue consumers can help to evenly distribute them across available queues.
This quickstart doesn't include any alerts. Do you think it should?
You can edit this quickstart to add helpful components. View the repository and open a pull request.