Skip to content

CAA records block certificate renewal after a hosting move

Trace a failed certificate renewal to CAA authorization, distinguish DNS errors from issuer restrictions, and verify the renewed public HTTPS endpoint.

By Crawlito · 6 min read

In short

  • CAA controls which certificate authorities may issue certificates for a domain.
  • A renewal can fail while the currently installed certificate still works.
  • Repair the confirmed authorization or DNS problem, then verify issuance and public deployment separately.
On this page

If certificate renewal fails after a hosting move, check whether the domain's CAA records authorize the certificate authority used by the new service. An existing certificate may continue working while a replacement cannot be issued. Confirm the issuer and the actual failure, correct the narrow DNS or authorization problem, and verify that the renewed certificate reaches the public endpoint.

Do not delete every CAA record as a first response. The records may express an intentional security policy shared by several services. The useful repair is the smallest change that matches the organization's approved certificate providers.

Separate today's working certificate from tomorrow's renewal

Record the renewal error, the hostname, the certificate provider, and the current certificate's expiry date. These describe different stages of the process.

A browser opening the page proves neither that renewal succeeded nor that the replacement certificate was deployed. Likewise, a failed renewal attempt does not necessarily mean the site is already unavailable.

Build a simple timeline: hosting change, certificate request, error, current certificate expiry, and next planned retry. This makes the urgency understandable without inventing a crawler outage that has not occurred.

The DNS and TLS guide explains why a future certificate failure can become an access problem before any page-level SEO rule is evaluated.

Check whether the failed request concerns a single hostname, a wildcard, or several names. The error should be tied to the names actually requested, rather than to the site's marketing name or a different domain in the account.

Read the issuer's explanation before editing DNS

Let's Encrypt's CAA documentation explains that CAA records restrict which authorities can issue certificates and that records can distinguish ordinary and wildcard issuance.

Your hosting provider may use another authority or manage multiple certificate products. Read the current provider documentation and the failed order details to identify the relevant authorization value.

Do not assume the company selling the hosting service is also the certificate issuer. A DNS entry naming the wrong organization will not authorize the intended authority.

Also distinguish an authorization denial from a failure to look up CAA records. A DNS timeout or validation problem can prevent the authority from establishing the policy even when the visible records appear correct.

Preserve the exact error message and timestamp. If the provider reports a different validation problem, such as domain-control verification, fix that problem instead of making unrelated CAA changes.

Inspect the relevant DNS records and their ownership

Query the affected name and review the provider's documented CAA lookup behavior:

bash
dig CAA example.com
dig CAA www.example.com

These commands inspect DNS; they do not establish the complete certificate authority's validation outcome by themselves. Record the resolver, response status, and returned records.

Check the authoritative DNS control panel as well. A registrar interface may display settings that are not served by the domain's current authoritative provider.

If a migration changed nameservers, confirm that the intended CAA records moved with the zone. The DNSSEC migration guide covers a separate reason validating DNS lookups can fail after such a move.

Keep related records and their purpose visible during review. A record that supports a separate email, application, or asset service may be legitimate even if it is unrelated to the current certificate request.

Do not equate an empty answer from one query with proof that no relevant restriction exists. Use the issuer's documented evaluation rules and error details to interpret the result.

Correct the approved issuer policy narrowly

For a site deliberately using Let's Encrypt, its documentation gives letsencrypt.org as the issuer identifier. A minimal illustrative record is:

example.com. CAA 0 issue "letsencrypt.org"

This is an example for that authority, not a universal record to paste into every domain. Confirm the actual provider before changing production DNS.

Review wildcard policy separately when the request includes a wildcard. The issuer's documentation distinguishes issue and issuewild records; blindly adding one record can leave the relevant restriction unchanged.

Preserve approved authorizations for other services. If the organization wants to retire an authority, treat that as a deliberate policy change with the certificate owners involved.

When the underlying issue is DNS lookup failure, repair resolution rather than broadening issuer permissions. An authorization change cannot compensate for a resolver being unable to obtain a trustworthy answer.

Save the previous record set and the intended replacement. That gives the DNS owner a concrete change to review and a reliable rollback reference if an unexpected dependency appears.

Retry through the certificate manager and inspect the result

Allow for the DNS update's actual publication and cache behavior, then retry issuance through the provider's supported workflow. Avoid repeated manual requests without understanding the previous error.

A successful order is one checkpoint. Confirm the requested names, issuer, and validity period of the certificate that was actually issued.

Next, verify deployment. A certificate manager can obtain a replacement while a load balancer or reverse proxy continues serving the old certificate.

Inspect the ordinary public endpoint and compare it with the newly issued certificate. If several endpoints terminate TLS, check the relevant pool rather than only one administrative console.

The certificate-chain guide covers the installation issue where the correct leaf certificate is present but the served chain is incomplete.

Keep issuance and deployment results separate in the completion note. This prevents a green certificate-order screen from being mistaken for proof that visitors receive the repaired configuration.

Test the content path without bypassing TLS

Request a representative public page using normal certificate verification. Confirm both successful TLS and the expected HTTP response.

If the domain redirects to another hostname, follow that path and verify the destination too. A healthy certificate on the first hostname does not establish that the full reader journey works.

Check separate content hosts that were included in the original incident, such as documentation or images. Their certificate lifecycles may be managed independently.

Do not ask visitors or crawlers to ignore certificate warnings. The repair should restore ordinary supported access.

A useful acceptance record includes the public hostname, certificate identity, successful verification, expected page title, and the observation time. Those details establish what was tested without claiming that every crawler has already revisited.

If a page still fails after TLS succeeds, continue with its application response and access rules. Keep that later defect distinct from the resolved certificate-issuance problem.

Make the next renewal review practical

Document the intended certificate authorities and the services that depend on them. Update that inventory during hosting migrations, CDN changes, and wildcard-certificate adoption.

Monitor the renewal process itself as well as public expiry. An early renewal failure is actionable before the current certificate expires.

For an agency handoff, identify who controls authoritative DNS, who manages certificate orders, and who deploys certificates to the public service. Different owners need a shared completion definition.

Do not report this repair as a proven SEO gain. The verified outcome is a working issuance and deployment path that preserves normal HTTPS access. Search performance needs separate measurement.

Keep the before error, DNS change, successful order, and public verification together. This evidence makes a future recurrence much easier to distinguish from an unrelated DNS or application incident.

Frequently asked questions

Can CAA break an already issued certificate immediately?

CAA governs issuance. A failed renewal can become an outage later when the currently deployed certificate expires or must be replaced.

Should I remove CAA entirely to make renewal work?

First identify the intended issuer and the exact failure. Preserve the approved policy and make a focused correction rather than deleting restrictions indiscriminately.

Is successful issuance enough to close the incident?

No. Confirm that the public endpoint serves the new certificate and that an ordinary verified request reaches the expected page.

After public HTTPS is healthy, run the free Crawlito check on the site's main landing page to inspect its current page-level readiness.

Filed under Access and robots.txt. Topics: dns, tls, website-migration