Console administration

Create and operate customer status pages

Set up customer components, publish service health, and pause monitoring for planned maintenance.

Version 1.0 Status Preview For Tenant administratorsFor Tenant ownersFor Responders status-pages monitoring maintenance

Status pages show customers the health of named components, such as Customer API or Scheduled backups. You choose which services feed each component and write customer updates separately. Internal incident titles, payloads, and timelines are not published automatically.

These controls are part of the status-page rollout. The console and mobile controls described here require a build containing them; CLI status commands require WarnFire client 1.1.0 or later. Private publication and invitations also require private hosting and email delivery to be enabled in your environment. The console shows which actions are available.

Before you begin

Use a workspace where the relevant services are configured. Workspace members can read status pages; workspace owners and administrators can create and manage them. The mobile app also requires an enabled responder profile associated with the same workspace account.

The page list shows your current allowance. Trial and Team plans allow three pages with 25 components per page; Business allows ten pages with 50 components per page. Complimentary billing preserves the assigned plan’s allowance and private-page eligibility until complimentary access expires. An explicit complimentary plan has the Team baseline. Enterprise defaults to 25 pages with 50 components per page, unless your active contract specifies other limits. HomeLab has no page allowance. Business and Enterprise allow private pages; publication also requires private hosting to be enabled.

Create a page and its components

Open Status pages in the workspace console sidebar.
Select Create status page, then enter a Page title and description, choose a timezone from the list and select visibility, then select Create page. The page title accepts up to 200 Unicode characters, including accented text and emoji. The logo URL is optional and must use HTTPS.
Open the page and add a component for each part of the service customers should see. Use the component’s move controls to set the display order.
Expand a component and add its Linked services. Choose the service, minimum incident severity, and debounce interval in seconds, then save the mapping.
Review Customer preview. Check the names, component states, and customer updates before publishing.

A component follows its enabled service mappings. The minimum severity determines which incidents qualify; debounce prevents a brief source state from immediately changing customer status. A monitor group is an organizational aid and does not itself publish a component. Heartbeats contribute through their owning service’s explicit mapping.

Publish, withdraw, and delete

For a public page, select Publish and confirm. The console shows its public URL when public hosting is configured. Publication is queued; open the URL after processing and check what customers can actually read. A preview inside the console does not prove that the public page is serving.

Withdraw removes public access asynchronously. Wait for the withdrawal state to finish and verify the former URL no longer serves the page. Use Delete page only when you intend to remove the page and its associated configuration, rather than temporarily hide it. Confirmations are required for these actions. If publishing is unavailable, check the page visibility and whether your environment has public hosting enabled.

Publish or withdraw through the API

Tenant API keys with status-pages:manage can call POST /v1/tenant/status-pages/{pageID}/publish and POST /v1/tenant/status-pages/{pageID}/unpublish, using Authorization: Bearer <API key> against https://api.warnfire.com. Read-only keys cannot use these operations. Each request is limited to the key’s workspace and recorded with that key’s audit identity.

These requests use the same plan, visibility, and hosting checks as the console. Inspect the returned publication_state: queued means publication is pending; purging means withdrawal is pending. A successful request does not prove the hosted copy has changed yet. Publication needs enabled hosting; withdrawal remains available when new publication is disabled.

See the API Explorer for the response contract.

Publish a private page and invite viewers

A private page has a separate sign-in address. Each page has its own generated identifier and hostname, including pages in the same workspace. Renaming a page does not change that identifier.

Choose Private visibility and review the customer preview.
Under Viewer access, enter the external viewer’s email and select Send invitation. The invitation grants read access to this page and creates no billable workspace seat. Workspace members can read the page using their existing WarnFire account.
Select Publish page and confirm. Wait for publication to finish, then open the private URL and sign in. A preview in the console does not prove the hosted page is serving.
An invited viewer opens the email link and selects Continue with invitation. Invitations expire after 24 hours and can be used once. An existing invited viewer can request another link from the page’s sign-in screen.

Select Remove access to revoke an external viewer. Removed access stops renewing immediately; access already issued to a browser expires within five minutes. Removing workspace membership, revoking its sign-in session, withdrawing the page, or losing the private-page plan allowance also blocks renewed access. A plan change never makes a private page public.

Re-invite restores a removed viewer with a new invitation. Resend invitation replaces an existing invitation. Old invitation proofs and sessions cannot become valid again. Delete viewer removes the retained viewer entry, pending invitations, proofs, and sessions; audit history remains. Deleting a viewer has the same five-minute bound for signed access already issued to a browser.

Private reading needs periodic access renewal. If renewal is temporarily unavailable, the page displays a warning; expired access requires signing in again. Access controls cannot recall content someone has already downloaded or copied. Withdraw or delete a private page to remove its hosted copy; wait for withdrawal to finish before changing its visibility.

Connect a custom private hostname

Custom private domains require Business or Enterprise and enabled custom-domain hosting in your environment. Open the private page, expand Custom private domains, and select Manage domains. Add the hostname you own, such as status.example.com, without a scheme or path.

Add the ownership TXT record shown in the console. Its name is _warnfire-status.<hostname> and its value begins warnfire-status=. You can use the displayed ownership CNAME alternative instead. Copy the exact value and keep the ownership record while connected.
Select Check DNS. After ownership is verified, the console immediately shows the routing record for your environment.
Point your new subdomain to the displayed WarnFire target using a CNAME: cdn.warnfire.com for production or cdn.dev.warnfire.com for development. If your DNS provider offers a proxy, use DNS-only routing. WarnFire handles HTTPS issuance and renewal automatically after routing; no separate ACM certificate-validation record is needed. Root domains require provider-specific setup; contact support before connecting one or moving a hostname that already serves traffic.
Select Check DNS and refresh until the domain is Active. Open the HTTPS hostname and verify sign-in, authorized reading, and denial for an unauthorized browser. DNS propagation and certificate issuance can take several minutes.

Custom hostnames preserve private-page access checks. The generated WarnFire hostname remains available. Removing a hostname is asynchronous: wait for removal to complete and verify the address no longer serves the page before reusing it. Public-page custom domains are not part of this private-domain workflow.

Pause a component for planned maintenance

Expand the component you intend to reboot or maintain.
Select Pause, choose a duration from 1 minute to 24 hours, and write an internal maintenance reason. Optionally add a separate customer message.
Select Pause monitoring. Only the customer message appears on the hosted page and in detailed subscriber emails; the internal reason stays in the console.
Check the recorded end time. The same button now reads Resume.
Select Resume when work finishes early, or let the window expire automatically.

Pause puts monitors belonging to the component’s enabled linked services into planned maintenance. This also affects those services if they are shared with other components. The status page stays online and the component shows maintenance. Existing incidents are preserved; pausing does not resolve them or erase earlier outages. Monitoring resumes when maintenance ends, with a fresh monitoring deadline rather than treating the planned pause as a missed check-in.

Service mappings cannot be changed while that component’s maintenance is active. Resume first, change the mappings, then pause again if needed. If several maintenance windows cover the same service, ending one leaves the other windows in effect.

The web dialog lets you choose from one minute through 24 hours and defaults to 30 minutes. Mobile Pause uses 30 minutes. The CLI status commands also accept --for within that range. Availability charts exclude planned maintenance from the uptime denominator.

Set a customer display override

Under Manual status override, choose a status, enter a reason, and optionally set an expiry in local time. Select Set override. Without an expiry, the override remains until you clear it.

An override changes the customer-facing component state; it does not pause monitors. Use Pause for a reboot. Use an override when the automatically derived state needs a deliberate customer-facing correction. Clear it when the normal state should apply again.

Write customer incident updates

Create an incident under Incident updates with a customer-facing title, initial update, incident state, and affected components. Post subsequent updates as the investigation progresses. Updates retain their recorded state and post time when their text is corrected; a resolved incident no longer offers a further-update form.

Write plain text intended for customers. Avoid credentials, internal payloads, private links, and details that are only appropriate in the internal incident timeline. Review the customer preview after posting.

Visitors choose Subscribe to updates on the hosted page, enter their email address, then open the confirmation email and select Confirm subscription. The console’s Email notifications panel also links to subscriber signup. Subscribers receive new customer incident updates and component status changes after those changes have been published. Publication refreshes alone do not send email.

Each workspace includes 100 verified email subscribers on Team, 500 on Business, and at least 1,000 on Enterprise, shared across public and private pages. An address subscribed to several pages counts once toward that limit. Consent remains page-specific: subscribing to one page does not subscribe you to another. Plan changes apply when a new subscriber confirms; existing subscribers are preserved if a workspace moves to a lower limit.

Private-page subscriptions require an enabled workspace membership or an accepted viewer invitation. Subscription confirmation never grants private-page access. Generic private emails announce an update and link to the page; readers must sign in to read the content. Owners and administrators can enable Include update details in emails to include the customer-facing title, state, and message. Delivered email content remains available to its recipient after page access is removed. Current access is checked again before sending. Withdrawn pages and revoked viewers stop receiving updates.

Status-page emails include a plain-text version and a branded HTML version with the original WarnFire flame logo used by the app. The recipient’s email client chooses which to display; both versions contain the same content and links. Detailed update emails include the customer incident or component name and state in the subject. Generic emails omit those details.

Every message includes Unsubscribe from this page. Open the link and confirm Unsubscribe to stop that page’s emails. Other page subscriptions remain active. Invitations to read a private page and subscriptions to receive updates are separate choices.

Workspace owners and administrators can open the Subscribers tab under Status pages to manage subscriptions across all lists. Search by email or filter by page. Each address shows a toggle for each page and its state: Verified, Pending confirmation, or Unsubscribed. Switching a toggle off removes that page’s subscription and cancels its pending emails and confirmation links. Switching it on requests email confirmation; it does not verify consent automatically.

Subscribe to all pages requests confirmation for each currently published page for which the address is eligible. Existing verified subscriptions are unchanged. Private pages still require existing membership or accepted viewing access. Unpublished pages are excluded, and pages created later are not automatically added. Repeat requests for pending subscriptions are throttled for 15 minutes.

Unsubscribe from all pages removes all that address’s subscriptions in the workspace. Delete subscriber explicitly removes the address and all its memberships after confirmation. Both cancel pending delivery and confirmation links and free the verified subscriber slot when no verified membership remains. An address still subscribed to another page continues to use one slot. Private-page viewing access is separate and remains unchanged. An address with no subscriptions can remain on screen until you leave the tab, allowing you to request confirmation again; it has no stored subscription or reserved slot. Load more subscribers continues through large lists.

Read times in the page’s timezone

The hosted page names its configured timezone and displays a live clock, its last-updated time, and incident update timestamps in that timezone. Daylight-saving changes are applied automatically. Email update timestamps use the same timezone. The clock synchronizes from the server; if synchronization is unavailable, the page labels the clock accordingly instead of silently trusting an incorrect device clock.

Use the mobile app

In a compatible WarnFire mobile build , open Pages, select a page, and read its components and updates. Each manageable component has one stateful Pause / Resume button: Pause takes a tap and confirmation; Resume takes one tap. Pause defaults to 30 minutes and expires automatically. Members without management access can read the page.

Page creation, service mappings, publishing, and customer update authoring use the web console.

Verify it worked

Check the selected workspace and page, review Customer preview, and verify the public URL after publication finishes. For planned maintenance, confirm the component shows maintenance with an end time, then confirm Resume or automatic expiry returns monitoring to its normal state. Verify shared services before assuming a component’s maintenance affects only that component.

If a save or pause request fails with an uncertain result, refresh before retrying. If a save succeeded but refresh failed, reload the page to read the saved state. Creation limits come from your current plan and active Enterprise contract, not from the number of components displayed in another page.

Next step

Use the WarnFire CLI reference to inspect pages or pause a whole page or component from a terminal.

The page table keeps every page visible and shows its publication state, published/current revisions, and last automatic publication time in the page’s timezone. Each row has page settings, publish, unpublish, and delete controls. Status and timestamps refresh every 15 seconds.

Choose an existing workspace service when adding a component to link it automatically, or create a custom component. Assign components to named groups; groups appear as top-level rows with disclosure triangles on the customer page. Group status reflects the most affected component.

Use Create incident for a new customer incident and Post incident update to follow up on an existing one. Incident titles allow up to 200 Unicode characters. Customer update bodies allow 4,000 UTF-8 bytes; the form shows the count before submission. Optional internal notes allow 1,000 UTF-8 bytes and remain in the console, excluded from hosted pages and subscriber email. Edit update corrects the customer message and its internal note, including clearing the internal note. Edit incident title uses the same 200-character limit. Edits preserve state and post time and do not resend delivered email. Delete update removes that update from subsequent publications and cancels unsent notifications. It does not change the incident’s state or recall delivered email.

Manual overrides also separate the internal reason from an optional customer message. The customer message appears with the component while the override is active. Clearing or expiring the override removes it. The internal reason is never published.

Team, Business, and Enterprise workspaces can hide the WarnFire attribution and free-workspace signup invitation in page settings. Other pages include the invitation below their incident history.

Availability history

Each reader shows 30- and 90-day timelines for every component and visual group. Operational time is green; degraded performance and partial outages are yellow; major outages are red; maintenance is purple; unknown time is gray. Hover over, focus, or tap a timeline interval to read its exact state, start/end times, page timezone and duration. Read interval history provides the same information as text, including without JavaScript. Brief outages have a minimum-width marker so they remain visible; the interval’s exact duration determines the calculation.

Uptime is operational time divided by known time excluding maintenance. Degraded performance and partial or major outages count as downtime. Unknown time is excluded from the denominator and never treated as healthy; coverage reports the portion of the window with recorded history. A window with no known non-maintenance time shows Unavailable, rather than 100% uptime. The chart states the publication time through which history was calculated.

Group history follows the membership that existed at each change and shows the worst member state. Overlapping member outages count once, and changing or removing a component does not rewrite the previous group’s history. Existing component state transitions are retained, but group membership before this history was introduced is unknown. Deleting a page or workspace removes its availability history. Internal notes and service mappings never appear in charts.

Select destinations from Services

On Services, choose Status page destinations beside the owning service. Select one or more customer components across public or private pages. Select additional source services to feed those same components as a service group. The component name is the customer-visible group name; its automatic status uses the worst active member and recovers after all contributing members recover. New mappings use error severity and a 3-minute debounce. Existing settings are preserved; adjust them under Status pages → Linked services. Component groups are visual groupings of components, separate from these groups of source services.

Resume maintenance before changing mappings. Saving several destinations makes individual mapping changes; if one fails, review the refreshed selections before retrying. No internal incident text is copied to a customer update.

Audited support preview

A platform owner managing a workspace can select View private page preview. This records an audit event and shows a static view of current customer content, including unpublished changes. Internal notes and scheduled future updates are excluded. The preview grants no hosted reader access and does not test sign-in, custom DNS, certificates, or the published copy. Test those through the hosted URL with an authorized reader account.