Aws secretsmanager additions

[mpechner-akasa]

This adds the ability to specify a secretKey in the awsSecretManager TriggerAuthentication. This will allow parsing of secrets that contain Key/Value pairs (returned in JSON format).

Resubmission of #6031

Provide a description of what has been changed

Checklist

• When introducing a new scaler, I agree with the scaling governance policy
• I have verified that my change is according to the deprecations & breaking changes policy
• Tests have been added
• Changelog has been updated and is aligned with our changelog requirements
• A PR is opened to update our Helm chart (repo) (if applicable, ie. when deployment manifests are modified)
• A PR is opened to update the documentation on (repo) (if applicable) Documented adding secretKey parameter to access aws secret manager keda-docs#1508
• Commits are signed with Developer Certificate of Origin (DCO - learn more)

Fixes #5940

Relates to #

[zroubalik]

Looking good, thanks. We should also update docs

[mpechner-akasa]

#5940

docs. kedacore/keda-docs#1508

[JorTurFer]

Really nice job! some comments inline

[zroubalik]

/run-e2e aws
Update: You can check the progress here

[mpechner-akasa]

/run-e2e aws Update: You can check the progress here

Will work on this next weekend 1/11-1/12

[JorTurFer]

Nice job! Thanks a lot ❤️

[JorTurFer]

/run-e2e aws
Update: You can check the progress here

[JorTurFer]

e2e tests are failing because the secret has the same name in both cases. This makes that one case passes and the other one fails because the e2e test is trying to create a secret with the same name and the secret is in deletion process

You can't create this secret because a secret with this name is already scheduled for deletion

You need to update the value of AwsCredentialsSecretName to include a different value per test case (as well as the random number to not conflict between tests). e.g: You can move the value calculation to the function getTemplateData and append a suffix with the test case

[JorTurFer]

I think that just using another secret name will be better as the test doesn't need to wait until the secret is deleted (not slowing the test execution). My suggestion was just appending the test case suffix to the secretManagerSecretName in the beginning of the test. I mean, currently the value of secretManagerSecretName is generated for both test in the startup of the file -> https://github.com/nrichardson-akasa/keda/blob/aws-secretsmanager-additions/tests/secret-providers/aws_secretmanager_pod_identity/aws_secretmanager_pod_identity_test.go#L62, the suggestion that I'm giving is overriding that variable during each case:

func TestAwsSecretManager(t *testing.T) {
	// Run the test twice with two different flag values
	flags := []bool{true, false}

	for _, useJSONSecretFormat := range flags {
		// THIS ASSIGNMENT SOLVES THE ISSUE
		secretManagerSecretName = fmt.Sprintf("connectionString-%d", GetRandomNumber())
		// Define a subtest for each flag value
		t.Run(getTestNameForFlag(useJSONSecretFormat), func(t *testing.T) {
			err := AwsSecretManager(t, useJSONSecretFormat)
			if err != nil {
				t.Errorf("AwsSecretManager(%v) failed: %v", getTestNameForFlag(useJSONSecretFormat), err)
			}
		})
	}
}

Just adding the line secretManagerSecretName = fmt.Sprintf("connectionString-%d", GetRandomNumber()) will use different secret names and you won't have to check if it has been deleted or not before creating it again