Amazon MQ for ActiveMQ Throughput benchmarks - Amazon MQ
mq.m4.largemq.m5.largemq.m5.xlargemq.m5.2xlargemq.m5.4xlarge
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Amazon MQ for ActiveMQ Throughput benchmarks

Benchmarking can help you choose the correct instance type and size for your workload messaging requirements. Scenarios for benchmarking include:

  • Cluster Stability: understanding how stable your cluster is during increasing, fluctuating, and stable load types.

  • Defining performance limits: approximating the maximum performance and throughput capabilities (i.e. cluster limits) of your cluster to help better scale your broker nodes when the number of messages published to your broker increases.

  • Optimal architecture and parameters: determining the most suitable architecture/parameters for your clusters, such as the number of destinations, persistent mode, message size, etc.

Amazon MQ provides benchmarking figures for the different instance types and sizes available for Amazon MQ for ActiveMQ.

Amazon MQ uses the ActiveMQ Maven 2 Performance Test to calculate benchmarks. Amazon MQ tests with an active/stanby deployment broker with an EFS volume as a storage type. The results are from a performance test conducted on a ECS cluster with a Fargate deployment which consists of a configuration of 4vCPUs and 16 GiBs of memory (equivalent to an EC2 m5.xlarge instance). Concurrent Store And Dispatch Queues (CSAD) are set to true for all tests performed. The duration for each test conducted is 5 minutes.

Note

When run in your own environment, results may differ by 3-6%.

The following tables provide performance and throughput benchmarks for Amazon MQ supported instance types to help you choose the correct instance sizes for your messaging workload.

mq.m4.large

Configuration options:

  • Broker Instance - mq.m4.large

  • Persistent - TRUE

  • Client - m5.xlarge

  • CSAD - TRUE

  • Protocol - Openwire

Producers/Consumers
Message size Metrics 25 50 100
1KB TPS 1849 3335 4665
CPU% 29% 37% 47%
5KB TPS 1672 2561 2970
CPU% 33% 47% 76%
10KB TPS 1586 1670 2268
CPU% 44% 87% 89%

mq.m5.large

Configuration options:

  • Broker Instance - mq.m5.large

  • Persistent - TRUE

  • Client - m5.xlarge

  • CSAD - TRUE

  • Protocol - Openwire

Producers/Consumers
Message size Metrics 25 50 100
1KB TPS 2247 4041 7566
CPU% 26% 32% 48%
5KB TPS 1636 3205 4443
CPU% 37% 63% 58%
10KB TPS 1668 3104 3227
CPU% 40% 53% 86%

mq.m5.xlarge

Configuration options:

  • Broker Instance - mq.m5.xlarge

  • Persistent - TRUE

  • Client - m5.xlarge

  • CSAD - TRUE

  • Protocol - Openwire

Producers/Consumers
Message size Metrics 25 50 100
1KB TPS 2255 3932 7453
CPU% 28% 32% 54%
5KB TPS 1766 3495 6215
CPU% 29% 51% 82%
10KB TPS 1641 3240 5613
CPU% 36% 61% 89%

mq.m5.2xlarge

Configuration options:

  • Broker Instance - mq.m5.2xlarge

  • Persistent - TRUE

  • Client - m5.xlarge

  • CSAD - TRUE

  • Protocol - Openwire

Producers/Consumers
Message size Metrics 25 50 100
1KB TPS 2025 4089 8093
CPU% 12% 18% 35%
5KB TPS 1865 3736 6845
CPU% 15% 27% 54%
10KB TPS 1747 3511 7057
CPU% 18% 36% 67%

mq.m5.4xlarge

Configuration options:

  • Broker Instance - mq.m5.4xlarge

  • Persistent - TRUE

  • Client - m5.xlarge

  • CSAD - TRUE

  • Protocol - Openwire

Producers/Consumers
Message size Metrics 25 50 100
1KB TPS 2094 4055 8153
CPU% 6% 9% 17%
5KB TPS 1742 3586 7158
CPU% 7% 13% 25%
10KB TPS 1733 3288 6671
CPU% 9% 16% 31%