Monitoring Amazon MQ brokers using Amazon CloudWatch - Amazon MQ
Logging and monitoring ActiveMQ brokersLogging and monitoring RabbitMQ brokers

Monitoring Amazon MQ brokers using Amazon CloudWatch

Amazon MQ and Amazon CloudWatch are integrated so you can use CloudWatch to view and analyze metrics for your ActiveMQ broker and the broker's destinations (queues and topics). You can view and analyze your Amazon MQ metrics from the CloudWatch console, the AWS CLI, or the CloudWatch CLI. CloudWatch metrics for Amazon MQ are automatically polled from the broker and then pushed to CloudWatch every minute.

For information, see Accessing CloudWatch metrics for Amazon MQ.

Note

The following statistics are valid for all of the metrics:

  • Average

  • Minimum

  • Maximum

  • Sum

The AWS/AmazonMQ namespace includes the following metrics.

Logging and monitoring ActiveMQ brokers

ActiveMQ broker metrics

Metric Unit Description
BurstBalance Percent The percentage of burst credits remaining on the Amazon EBS volume used to persist message data for throughput-optimized brokers. If this balance reaches zero, the IOPS provided by the Amazon EBS volume will decrease until the Burst Balance refills. For more information on how Burst Balances work in Amazon EBS, see: I/O Credits and Burst Performance.
CpuCreditBalance Credits (vCPU-minutes)
Important

This metric is available only for the mq.t2.micro broker instance type.

CPU credit metrics are available only at five-minute intervals.

The number of earned CPU credits that an instance has accrued since it was launched or started (including the number of launch credits). The credit balance is available for the broker instance to spend on bursts beyond the baseline CPU utilization.

Credits are accrued in the credit balance after they're earned and removed from the credit balance after they're spent. The credit balance has a maximum limit. Once the limit is reached, any newly earned credits are discarded.

CpuUtilization Percent The percentage of allocated Amazon EC2 compute units that the broker currently uses.
CurrentConnectionsCount Count The current number of active connections on the current broker.
EstablishedConnectionsCount Count The total number of connections, active and inactive, that have been established on the broker.
HeapUsage Percent The percentage of the ActiveMQ JVM memory limit that the broker currently uses.
InactiveDurableTopicSubscribersCount Count The number of inactive durable topic subscribers, up to a maximum of 2000.
JobSchedulerStorePercentUsage Percent The percentage of disk space used by the job scheduler store.
JournalFilesForFastRecovery Count The number of journal files that will be replayed after a clean shutdown.
JournalFilesForFullRecovery Count The number of journal files that will be replayed after an unclean shutdown.
NetworkIn Bytes The volume of incoming traffic for the broker.
NetworkOut Bytes The volume of outgoing traffic for the broker.
OpenTransactionCount Count The total number of transactions in progress.
StorePercentUsage Percent The percent used by the storage limit. If this reaches 100, the broker will refuse messages.
TempPercentUsage Percent The percentage of available temporary storage used by non-persistent messages.
TotalConsumerCount Count The number of message consumers subscribed to destinations on the current broker.
TotalMessageCount Count The number of messages stored on the broker.
TotalProducerCount Count The number of message producers active on destinations on the current broker.
VolumeReadOps Count The number of read operations performed on the Amazon EBS volume.
VolumeWriteOps Count The number of write operations performed on the Amazon EBS volume.

Dimensions for ActiveMQ broker metrics

Dimension Description
Broker

The name of the broker

Note

A single-instance broker has the suffix -1. An active/standby broker for high availability has the suffixes -1 and -2 for its redundant pair.

ActiveMQ destination (queue and topic) metrics

Important

The following metrics include per-minute counts for the CloudWatch polling period.

  • EnqueueCount

  • ExpiredCount

  • DequeueCount

  • DispatchCount

  • InFlightCount

For example, in a five-minute CloudWatch period, EnqueueCount has five count values, each for a one-minute portion of the period. The Minimum and Maximum statistics provide the lowest and highest per-minute value during the specified period.

Metric Unit Description
ConsumerCount Count The number of consumers subscribed to the destination.
EnqueueCount Count The number of messages sent to the destination, per minute.
EnqueueTime Time (milliseconds) The end-to-end latency from when a message arrives at a broker until it is delivered to a consumer.
Note

EnqueueTime does not measure the end-to-end latency from when a message is sent by a producer until it reaches the broker, nor the latency from when a message is received by a broker until it is acknowledged by the broker. Rather, EnqueueTime is the number of milliseconds from the moment a message is received by the broker until it is successfully delivered to a consumer.

ExpiredCount Count The number of messages that couldn't be delivered because they expired, per minute.
DispatchCount Count The number of messages sent to consumers, per minute.
DequeueCount Count The number of messages acknowledged by consumers, per minute.
InFlightCount Count The number of messages sent to consumers that have not been acknowledged.
ReceiveCount Count The number of messages that have been received from the remote broker for a duplex network connector.
MemoryUsage Percent The percentage of the memory limit that the destination currently uses.
ProducerCount Count The number of producers for the destination.
QueueSize Count The number of messages in the queue.
Important

This metric applies only to queues.
TotalEnqueueCount Count The total number of messages that have been sent to the broker.
TotalDequeueCount Count The total number of messages that have been consumed by clients.
Note

TotalEnqueueCount and TotalDequeueCount metrics include messages for advisory topics. For more information about advisory topic messages, see the ActiveMQ documentation.

Dimensions for ActiveMQ destination (queue and topic) metrics

Dimension Description
Broker

The name of the broker.

Note

A single-instance broker has the suffix -1. An active/standby broker for high availability has the suffixes -1 and -2 for its redundant pair.

Topic or Queue The name of the topic or queue.
NetworkConnector The name of the network connector.

Logging and monitoring RabbitMQ brokers

RabbitMQ broker metrics

Metric Unit Description
ExchangeCount Count The total number of exchanges configured on the broker.
QueueCount Count The total number of queues configured on the broker.
ConnectionCount Count The total number of connections established on the broker.
ChannelCount Count The total number of channels established on the broker.
ConsumerCount Count The total number of consumers connected to the broker.
MessageCount Count The total number of messages in the queues.
Note

The number produced is the total sum of ready and unackknowledged messages on the broker.

MessageReadyCount Count The total number of ready messages in the queues.
MessageUnacknowledgedCount Count The total number of unacknowledged messages in the queues.
PublishRate Count The rate at which messages are published to the broker.

The number produced represents the number of messages per second at the time of sampling.
ConfirmRate Count The rate at which the RabbitMQ server is confirming published messages. You can compare this metric with PublishRate to better understand how your broker is performing.

The number produced represents the number of messages per second at the time of sampling.
AckRate Count The rate at which messages are being acknowledged by consumers.

The number produced represents the number of messages per second at the time of sampling.

Dimensions for RabbitMQ broker metrics

Dimension Description
Broker

The name of the broker.

RabbitMQ node metrics

Metric Unit Description
SystemCpuUtilization Percent The percentage of allocated Amazon EC2 compute units that the broker currently uses.
RabbitMQMemLimit Bytes The RAM limit for a RabbitMQ node.
RabbitMQMemUsed Bytes The volume of RAM used by a RabbitMQ node. When memory use goes above the limit, the cluster will block all producer connections.
RabbitMQDiskFreeLimit Bytes The disk limit for a RabbitMQ node. This metric is different per instance size. For more information about Amazon MQ instance types, see
RabbitMQDiskFree Bytes The total volume of free disk space available in a RabbitMQ node. When disk usage goes above its limit, the cluster will block all producer connections.
RabbitMQFdUsed Count Number of file descriptors used.

Dimensions for RabbitMQ node metrics

Dimension Description
Node The name of the node.
Note

A node name consists of two parts: a prefix (usuallly rabbit) and a hostname. For example, rabbit@ip-10-0-0-230.us-west-2.compute.internal is a node name with the prefix rabbit and the hostname ip-10-0-0-230.us-west-2.compute.internal.

Broker

The name of the broker.

RabbitMQ queue metrics

Metric Unit Description
ConsumerCount Count The number of consumers subscribed to the queue.
MessageReadyCount Counter The number of messages that are currently available to be delivered.
MessageUnacknowledgedCount Count The number of messages for which the server is awaiting acknowledgement.
MessageCount Counter The total number of MessageReadyCount and MessageUnacknowledgedCount (also known as queue depth).

Dimensions for RabbitMQ queue metrics

Note

Amazon MQ for RabbitMQ will not publish metrics for virtual hosts and queues with names containing blank spaces, tabs or other non-ASCII characters.

For more information about dimension names, see Dimension in the Amazon CloudWatch API Reference.

Dimension Description
Queue The name of the queue.
Virtual host Name of the virtual host.