/Azure-IoTHub-StressTest

Azure IoT Hub Stress test tool using azure batch service

Primary LanguageC#

Azure-IoTHub-StressTest

A GUI tool to simulate millions of IoT devices to test the capability of the Azure IoT Hub. The simulated devices are running in Azure Batch service.

User Guidance

1. Set expectations for stress test

This tab helps to convert from expections, such as total device count, message rate, to the azure resources we need to run this test.

  • Tab1 Image

2.1 Provision Azure resources

Here is the list of azure resources and credentials we need to run the test:

  • Azure IoT Hub: owner connection string and Event Hub endpoint
  • Azure Batch: service URL and access key
  • Azure Storage Account: connection string (needed for batch service)

Hint:

  • You can also put credentials into app.config.
  • You should create more partitions for the IoT Hub. It will give better monitor performance.

3. Start running test

After you filled out all fields, the button "Start Stress Load Test" will be enabled. Click to see the progress and status.

If you see this , you can go ahead to the third tab to start monitoring stress test.

  • Tab1 Image

Hint:

  • If the test is already running, you can monitor it by choosing an existing batch job from the drop down list.
  • Tab1 Image

4.Monitor the test

We will start to pull message for a given partition to measure the performance. It may take a while for the connection to be established.

Here is what the monitoring screen looks like

  • Tab1 Image

Test result section:

  • Batch

    • Job Id: This is the job id on Azure Batch Service. You can use it to find the job on Azure Portal.
    • Task Status: A job consists of several tasks. This field indicates the status of all tasks that belongs to the job.
    • Job running time: It shows how long the test job has been running on cloud.

  • Last Fetched Message

    • From Device: Indicates the device from which we received the latest message.
    • Content: This is the latest message fetched from IoT Hub.

  • Details

    • Monitor running time: How long since local monitoring thread started.
    • Device-To-Hub Delay (average): The average time used for a message to reach IoT Hub (calculated from all received messages).
    • Device-To-Hub Delay (last minute): The average time used for a message to reach IoT Hub (calculated from last one minute messages).
    • Throughput for partition (last minute): The number of messages fetched by monitoring thread.
    • Throughput for IoT Hub (last minute): Estimated total throughput.