fix(godaddy): Handle missing Retry-After header gracefully

[alexstojda]

Description

Based on some testing, it seems GoDaddy has an undocumented, per-endpoint quota.

During startup, the client calls the endpoint https://api.godaddy.com/v1/domains?statuses=ACTIVE,PENDING_DNS_ACTIVE1 to test and validate that the credentials are working. Calling that endpoint with the same API key manually, I was able to reproduce the 429 response the API sends, which did not include the Retry-After header and thus caused the panic on startup.

However, I was able to call other endpoints, ex: https://api.godaddy.com/v1/domains/<domain>/records, leading me to believe there is an undocumented, per-endpoint quota.

With this in mind, I updated the request handler to gracefully fail without panicing when the header is missing, so that other requests can still be attempted.

Fixes #4864

Checklist

• Unit tests updated
• [N/A] End user documentation updated

[linux-foundation-easycla]

The committers listed above are authorized under a signed CLA.

• ✅ login: alexstojda / name: Alex (bd9aaa6, 8ab04cf, fc000f0, 31abb6f)

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by:
Once this PR has been reviewed and has the lgtm label, please assign raffo for approval. For more information see the Kubernetes Code Review Process.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[k8s-ci-robot]

Welcome @alexstojda!

It looks like this is your first PR to kubernetes-sigs/external-dns 🎉. Please refer to our pull request process documentation to help your PR have a smooth ride to approval.

You will be prompted by a bot to use commands during the review process. Do not be afraid to follow the prompts! It is okay to experiment. Here is the bot commands documentation.

You can also check if kubernetes-sigs/external-dns has its own contribution guidelines.

You may want to refer to our testing guide if you run into trouble with your tests not passing.

If you are having difficulty getting your pull request seen, please follow the recommended escalation practices. Also, for tips and tricks in the contribution process you may want to read the Kubernetes contributor cheat sheet. We want to make sure your contribution gets all the attention it needs!

Thank you, and welcome to Kubernetes. 😃

[k8s-ci-robot]

Hi @alexstojda. Thanks for your PR.

I'm waiting for a kubernetes-sigs member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[rayshoo]

Hi alextojda! I saw your comment and am writing here to discuss. Thank you for actively working to solve the same problem.

I also thought about reading the body message of the response to determine the exact situation of monthly quota exceeded.
The parent function CallApiWithContext function that calls Client.Do() appears to call UnmarshalResponse, which is also a function that reads the body message of the response, in the return value.

func (c *Client) CallAPIWithContext(ctx context.Context, method, path string, reqBody, resType interface{}, needAuth bool) error {
	req, err := c.NewRequest(method, path, reqBody, needAuth)
	if err != nil {
		return err
	}
	req = req.WithContext(ctx)
	response, err := c.Do(req)
	if err != nil {
		return err
	}
	return c.UnmarshalResponse(response, resType) // read body message with Reader
}

As a result, it was judged that reading the body message with io.Reader in Client.Do would empty the stream and cause problems in the subsequent function, UnmarshalResponse, so a method to bypass that method was used on PR #4873

In the case of a 429 response within the Do function, I think that to solve this problem by reading the body message, we need to create a new ReadCloser, read it, and fill the stream again.

[alexstojda]

@rayshoo thanks for pointing that out. After doing some testing, I decided a simpler approach would be to just handle the missing Retry-After header itself gracefully, and let the existing error-handling logic do the rest.

I also modified the error handling of the validation step on startup to only panic when the error is not a QUOTA_EXCEEDED.

[mloiseleur]

/ok-to-test

[mloiseleur]

@alexstojda BTW, you may have noticed that we have started the process to move providers out of tree.

Would you be interested in moving this provider out of tree ?

[mloiseleur]

Many thanks for adding a test.You'll need to add the license header and I should be able to lgtm it.

[alexstojda]

@alexstojda BTW, you may have noticed that we have started the process to move providers out of tree.

Would you be interested in moving this provider out of tree ?

@mloiseleur I could be interested in taking that on, but I predict I won't have the bandwidth to do so until April/May-ish.

Many thanks for adding a test.You'll need to add the license header and I should be able to lgtm it.

Done!

[mloiseleur]

@alexstojda Would you please detail why you are removing the needAuth boolean which is set to true on all call ?

[alexstojda]

My local linter complained about this as the variable is unused in the function.