Writing a Python canary script - Amazon CloudWatch

Writing a Python canary script

This script passes as a successful run, and returns a string. To see what a failing canary looks like, change fail = False to fail = True

def basic_custom_script(): # Insert your code here # Perform multi-step pass/fail check # Log decisions made and results to /tmp # Be sure to wait for all your code paths to complete # before returning control back to Synthetics. # In that way, your canary will not finish and report success # before your code has finished executing fail = False if fail: raise Exception("Failed basicCanary check.") return "Successfully completed basicCanary checks." def handler(event, context): return basic_custom_script()

Packaging your Python canary files

If you have more than one .py file or your script has a dependency, you can bundle them all into a single ZIP file. If you use the syn-python-selenium-1.1 runtime, the ZIP file must contain your main canary .py file within a python folder, such as python/my_canary_filename.py. If you use syn-python-selenium-1.1 or later, you can optionally use a different folder , such as python/myFolder/my_canary_filename.py.

This ZIP file should contain all necessary folders and files, but the other files do not need to be in the python folder.

Be sure to set your canary’s script entry point as my_canary_filename.functionName to match the file name and function name of your script’s entry point. If you are using the syn-python-selenium-1.0 runtime, then functionName must be handler. If you are using syn-python-selenium-1.1 or later, this handler name restriction doesn't apply, and you can also optionally store the canary in a separate folder such as python/myFolder/my_canary_filename.py. If you store it in a separate folder, specify that path in your script entry point, such as myFolder/my_canary_filename.functionName.

Changing an existing Selenium script to use a Synthetics canary

You can quickly modify an existing script for Python and Selenium to be used as a canary. For more information about Selenium, see www.selenium.dev/.

For this example, we'll start with the following Selenium script:

from selenium import webdriver def basic_selenium_script(): browser = webdriver.Chrome() browser.get('https://example.com') browser.save_screenshot('loaded.png') basic_selenium_script()

The conversion steps are as follows.

To convert a Selenium script to be used as a canary
  1. Change the import statement to use Selenium from the aws_synthetics module:

    from aws_synthetics.selenium import synthetics_webdriver as webdriver

    The Selenium module from aws_synthetics ensures that the canary can emit metrics and logs, generate a HAR file, and work with other CloudWatch Synthetics features.

  2. Create a handler function and call your Selenium method. The handler is the entry point function for the script.

    If you are using syn-python-selenium-1.0, the handler function must be named handler. If you are using syn-python-selenium-1.1 or later, the function can have any name, but it must be the same name that is used in the script. Also, if you are using syn-python-selenium-1.1 or later, you can store your scripts under any folder and specify that folder as part of the handler name.

    def handler(event, context): basic_selenium_script()

The script is now updated to be a CloudWatch Synthetics canary. Here is the updated script:

The webdriver is an instance of the class SyntheticsWebDriver and the browser returned by webdriver.Chrome() is an instance of SyntheticsBrowser.

from aws_synthetics.selenium import synthetics_webdriver as webdriver def basic_selenium_script(): browser = webdriver.Chrome() browser.get('https://example.com') browser.save_screenshot('loaded.png') def handler(event, context): basic_selenium_script()

Changing an existing Puppeteer Synthetics script to authenticate non-standard certificates

One important use case for Synthetics canaries is for you to monitor your own endpoints. If you want to monitor an endpoint that isn't ready for external traffic, this monitoring can sometimes mean that you don't have a proper certificate signed by a trusted third-party certificate authority.

Two possible solutions to this scenario are as follows:

You are not limited to these two options when you use CloudWatch Synthetics canaries. You can extend these features and add your business logic by extending the canary code.

Note

Synthetics canaries running on Python runtimes innately have the --ignore-certificate-errors flag enabled, so those canaries shouldn't have any issues reaching sites with non-standard certificate configurations.