Help center Getting started Administration Keep your customer portal secure with managed HTTPS and troubleshoot certificate problems Updated Wednesday, October 28, 2026 · Marcel Gravelle HTTPS protects the connection between a customer’s browser and your support portal. For a branded portal connected through Core’s Custom domain workflow, Supportbench provisions a managed TLS certificate through Azure Front Door. Your team maintains the domain and verifies the customer experience; it does not generate a certificate signing request, buy a certificate or upload a PFX in this Core workflow. This guide explains what to check during setup and when a working address develops a certificate or access problem. At a glanceFor: support operations leads, portal administrators and the DNS or identity owners supporting a branded customer portal. Applies to: Supportbench Core.Start here: Settings → Self-Service → Customer portal → select the portal → Portal address → Custom domain Understand the three checks a customer connection must pass · What Core manages, and what your team still owns · Follow the full domain procedure for a new address · Read the status before choosing an action · Respond to a browser certificate warning · Maintain the domain after launch · When HTTPS works but the portal still does not · Move an existing Legacy certificate setup carefully · Prepare a useful Supportbench escalation · Example: a portal issue during a weekly account review Understand the three checks a customer connection must pass LayerWhat it doesWhat success does not proveDNS and hostnameDirect the customer’s chosen address to the intended service. A direct CNAME record is part of this managed setup.Correct DNS alone does not mean the managed certificate is ready.HTTPS / TLS certificateLets the browser establish an encrypted connection and validate the hostname’s certificate. SSL is the older term often used for the same operational topic.A valid certificate does not authenticate a particular customer or authorize a case or article.Sign-in and customer accessIdentify the customer and enforce the applicable portal, Company/Contact and knowledge restrictions. SAML is one possible sign-in method.Successful sign-in does not mean every article or every case should be visible.Keep these layers separate during diagnosis. A browser certificate warning happens before the normal customer workflow is trustworthy. A portal that opens securely but rejects SAML has a different problem. A signed-in contact who cannot see a restricted article may be encountering the intended access policy.How this helps B2B operationsA strategic customer may have dozens of contacts who all rely on the same portal. Identify which layer is failing before asking each contact to reset a password or before changing a shared access policy. One clear incident record with the hostname, status and failing step helps DNS, identity and support teams coordinate an effective response. What Core manages, and what your team still owns Core’s Custom domain workflow requests an Azure-managed certificate for the approved hostname. The current form has no certificate-file upload, private-key field, CSR generator, expiry editor or manual Renew certificate button. Instructions that ask you to export a PFX and upload it describe an older setup; do not use them as the procedure for a new managed Core domain.OwnerResponsibilitySupportbench managed workflowCoordinate domain validation, managed TLS provisioning, portal publication and subsequent domain-state checks.Your DNS ownerControl the domain, add the exact requested records, preserve the direct CNAME and review later DNS changes for impact.Your identity ownerKeep SAML provider values aligned with the actual portal hostname where SAML is used.Your support operations ownerTest the customer journey, maintain useful support links, record who owns each dependency and coordinate customer communication.Azure automatically renews eligible managed certificates; a direct CNAME to the Front Door endpoint is a renewal requirement. Core periodically rechecks the managed state and can show Action required when TLS revalidation is needed. This is managed operation, not a promise that DNS can be changed arbitrarily or that a domain will remain healthy without oversight. See Microsoft’s official Azure Front Door domain and certificate guidance for the underlying platform behavior. Use the Supportbench controls below for your portal; Azure administration examples are not additional buttons in Core. Follow the full domain procedure for a new address Confirm the intended portal and an organization-controlled subdomain. You need permission to edit that portal, and managed custom domains must be enabled for the environment.Open Settings → Self-Service → Customer portal, select the portal, choose Portal address, and find Custom domain.Follow the complete custom-domain setup procedure, including SAML coordination when enabled, the separate Request custom domain action, and the exact displayed TXT and direct CNAME records.Wait for Active and protected by managed TLS. Do not distribute the new address while the certificate is still provisioning.Test the secure connection, sign-in, allowed content and excluded content as separate checks. Record the tested hostname, portal and date.A TXT record proves domain control during validation. The CNAME directs traffic and is important for managed renewal. Neither record grants customer access. Copy the names and values displayed for your own request; do not reuse an example hostname or another account’s record. Read the status before choosing an action Open the same Custom domain card and compare the exact status with the table. These are Supportbench labels; the Azure portal may describe its underlying states differently.Status in CoreWhat to doPreparing DNS validationWait for request-specific DNS instructions. Repeated removal and recreation is not a shortcut.Add the TXT record below to verify ownershipHave the DNS owner compare the complete name and exact value with the displayed request. Check the correct DNS zone and whether the provider appends the domain automatically.Point the direct CNAME below to SupportbenchAdd or correct the exact displayed direct target. Review conflicting records and avoid proxying or flattening that hides the required CNAME.Azure Front Door is issuing the TLS certificateAllow provisioning to progress. A successful DNS check is not permission to ignore a browser certificate warning.Active and protected by managed TLSPerform or repeat the customer workflow checks. If the browser still warns, confirm the exact address and investigate the discrepancy.Action is required before this domain can continueReview the hostname and current DNS with its owner. Core can reach this state when later TLS revalidation is required. Use Retry after correcting a confirmed problem when offered; escalate a persistent state.The domain request needs attentionReview the visible status or error, exact hostname and DNS evidence. Correct a confirmed problem and use Retry if shown. Detailed underlying failure text may require Supportbench investigation.Removing the managed domain / RemovedVerify whether removal was intentional and which alternative route customers should use. Removal is a separate managed process, not a temporary certificate refresh.Core does not promise a fixed issuance or recovery time in this workflow. Capture when you checked and what changed. A provider DNS outage, an incorrect record and a certificate-issuance failure need different investigation; the status alone is not a complete diagnosis. Respond to a browser certificate warning Stop before entering customer credentials or sensitive information at the warning page. Do not tell customers to bypass the browser warning as a permanent workaround.Record the exact hostname and the browser’s error text without capturing private case content, passwords or tokens. Check for a misspelling or an old bookmarked hostname.Open the authorized Supportbench settings through a known working route and inspect the managed-domain status for the intended portal. Confirm whether the hostname is current, still provisioning, being replaced or removed.Ask the DNS owner to compare the current direct CNAME with the target shown by Core. Include recent DNS, proxy or hosting changes in the investigation. Do not alter another application’s record to test a guess.If Active is shown but the warning persists, have the appropriate IT or Supportbench owner investigate the certificate and routing discrepancy. Record whether the problem occurs for more than one approved test browser or network; an isolated local inspection or device problem is different from a general portal outage.Use an already verified alternative support channel or portal entry point under your incident process. Explain the next update time to affected customers. Do not improvise a public workaround that exposes restricted support records.After correction, retest HTTPS and the complete customer workflow before closing the incident. Maintain the domain after launch Keep the direct CNAME aligned with the managed endpoint and tell the support owner before moving the hostname to a proxy, redirect service, different host or DNS provider. A change that appears unrelated to Supportbench can affect certificate validation or renewal. The TXT record used for an earlier request may not satisfy a later request if a new validation value is issued. Follow the current instructions and coordinate with Supportbench when the required action is unclear.Record the portal, hostname, DNS zone owner, identity owner and a tested alternative support route. Review this information when staff responsibilities change.Include the portal in your existing service checks. Test a real customer entry path and sign-in separately from the homepage. This recommendation does not imply that Core supplies a new external monitoring or certificate-expiry alert feature.Review customer-facing links after a domain replacement. Do not assume the old address will redirect indefinitely.When a later Action required or Failed state appears, investigate the current request and DNS before using Retry. Retry does not repair a DNS record or update an identity provider.If organizational policy requires a particular certificate issuer or private-key arrangement, discuss compatibility with Supportbench before changing the domain. The Core managed form does not offer a customer-managed certificate upload.Managed certificates can change during renewal. Avoid building a customer process around a fixed certificate fingerprint. If an enterprise customer reports a certificate-pinning or gateway-policy issue, involve its IT owner and Supportbench rather than instructing end users to disable validation. When HTTPS works but the portal still does not Observed resultNext investigationSecure portal opens, SAML sign-in failsCheck the provider’s ACS and Entity ID against the values shown for the current hostname. A certificate renewal does not fix a mismatched sign-in destination.Sign-in succeeds but an article is unavailableCheck the article’s publication and audience, the portal’s Topics and Divisions, and the customer’s authorized access. HTTPS does not publish knowledge.A contact sees the wrong set of casesReview Company/Contact relationships and customer visibility rules. Do not treat changing the hostname or certificate as an access correction.The old bookmark fails after replacementVerify the new approved hostname and update the relevant customer-facing links. Replacement does not promise a permanent old-domain redirect.The general settings Save Changes did not retry TLSUse the dedicated Custom domain action. Retry appears for applicable failed or action-required states; it is separate from saving portal fields.The card or edit action is missingCheck portal edit permissions and environment availability with an administrator. Do not switch to the obsolete PFX recipe merely because this account lacks the current control. Move an existing Legacy certificate setup carefully If your portal currently relies on a Legacy certificate upload or an older custom-domain binding, first identify its owner and working address. Keep the existing binding and certificate material protected while you coordinate a migration with Supportbench, the DNS owner and the SAML owner. The old certificate may still be serving real customers until the replacement is ready.Do not delete an old private key, remove its binding or change a live CNAME as the first migration step. Agree the new hostname or migration route, complete managed validation, and verify customer access before retiring the old setup under the owner’s change process. This guide replaces the old setup instructions for current Core; it does not authorize removal of existing infrastructure or private keys. Prepare a useful Supportbench escalation When the issue remains after checking the intended hostname and exact DNS records, provide evidence that lets the right team reproduce the problem:Portal name, current and requested hostname, and whether this is initial setup, replacement, renewal investigation or removal.Exact visible status, browser error text and time observed, with the time zone. Include a screenshot of only the relevant controls if it contains no private customer information.The DNS owner’s confirmation of the public TXT/CNAME names and the CNAME target, plus recent DNS or proxy changes. Share validation details only through the approved support channel.Whether the HTTPS homepage opens, whether SAML completes, and which specific permitted customer action fails. Keep these results separate.Which approved test browsers or networks were checked, and whether other users are affected.The existing support case and responsible DNS/identity contacts, so teams can coordinate without starting parallel untracked changes.Do not attach certificate private keys, PFX passwords, administrator passwords, authentication cookies or sign-in tokens. The current Core managed setup does not need you to send private certificate material. Example: a portal issue during a weekly account review An account manager hears that several contacts cannot use the usual support bookmark. The support lead first confirms the exact hostname: it is the address replaced last week, while the new managed domain is Active. A test contact can sign in at the new address and open the intended account’s cases. The team updates the customer’s onboarding resource and shared links, records the result in the support case, and confirms that the old hostname is no longer the agreed entry point.In a different incident, the current hostname presents a certificate warning after a DNS provider change. The same lead records the warning and managed status, asks the DNS owner to compare the direct CNAME, and gives the customer a verified support channel while the responsible teams investigate. Keeping the symptom, evidence and owner explicit avoids an unnecessary access-policy change or password reset. Continue learning Set up, replace or remove a branded portal domainConfigure the customer portal and its audienceConfigure customer-portal SAMLPlan product-specific knowledge portals Was this article helpful? Yes No Thank you for your feedback.