Skip to main content
Network

Web Content Change

Detect unexpected changes to the content of your web pages, or pages that have stopped updating.

The Web Content Change check fetches one or more web pages and reports when their content differs from the last accepted reference. This lets you notice unexpected changes to a website, such as tampering. In reverse mode it instead reports a page that stops changing, for example a status page that is no longer updated. The check runs on Windows, Linux, and macOS (x64 and Arm64 each) every 60 minutes. The interval belongs to the check itself and cannot be changed per assignment.

Web Content Change or HTTP Content Check?

GoalMatching check
Detect any change to a page without deciding in advance which text might changeWeb Content Change
Detect that a page has stopped updatingWeb Content Change with Alert when content is unchanged
Check whether a specific text is present or absent, or monitor status code and response timeHTTP Content Check
Monitor pages with constantly changing parts such as timestamps, ads or session tokensHTTP Content Check
Monitor endpoints that need authentication, custom headers or POST requestsHTTP Content Check

Web Content Change compares the complete response byte for byte. If even a single character changes on every request, the check reports a change on every run. For such pages, the HTTP Content Check with a carefully chosen text is the better fit.

Prerequisites

  • You need the Monitoring Check Management permission.
  • The target device must have a connected agent with network access to the web pages.

The requests originate from the device the check is assigned to. Every URL must therefore be reachable from there; this also lets you monitor internal pages that are not published on the internet.

Steps

  1. Go to Devices and open the device.
  2. Click the Checks tab → Add Check → Web Content Change.
  3. Under URLs, enter one or more complete addresses, for example https://www.example.com/imprint.
  4. Optional: adjust the remaining settings, such as Alert when content is unchanged for a status page.
  5. Click Add Check.

On the first run, the check stores the current content of each URL as the reference and reports OK. The detailed result text then shows "References established:" with the number of URLs. From the next run on, the check compares against this reference.

Accept a change

In normal mode, the check does not update the reference on its own. A detected change therefore stays Critical until you accept it. If the page returns to the reference content before that, the check reports OK again.

  1. Check whether the change to the page is intended.
  2. Open the check in the device's Checks tab, enable Accept current content and save the change.
  3. On the next run, the check stores the current content of each URL as the new reference and reports OK. URLs that return an error on that run keep their previous reference.
  4. Then disable Accept current content again.

Warning: While Accept current content is enabled, every run stores the current content as the new reference without review. The pages are still fetched, but changes do not raise an alert.

Settings

FieldTypeRequiredDefaultDescription
URLsListYes—Absolute http or https addresses of the pages to watch. The check fetches duplicate entries only once
Alert when content is unchangedbooleanNofalseReverse mode: alerts while the page stops changing
Accept current contentbooleanNofalseStores the current content as the new reference and reports success (see above)
Ignore certificate errorsbooleanNofalseAccepts any TLS certificate, including expired or self-signed ones
Timeout (ms)IntegerNo30000Timeout per request in milliseconds. Values below 1000 count as 1000, values above 120000 as 120000

How the check compares

  • The check fetches each URL with GET without additional headers and follows up to 5 redirects. It compares the complete content of the final response.
  • Only a SHA-256 hash of the content is stored, not the content itself. The check therefore cannot show you what changed.
  • If a server responds with a status code outside 2xx, the check treats this as an error, not as a content change. The reference stays unchanged.
  • In reverse mode (Alert when content is unchanged), the check reports OK as soon as the content has changed since the last run and stores it as the new reference. If the content stayed the same, it reports Critical. The page must therefore change between two runs, that is within 60 minutes.
  • The agent stores the references on the device, per URL and separately for each mode. If you switch the mode, the first run afterwards establishes the references again.

Results

Each URL is evaluated on its own; the overall result is the worst URL result.

SituationResult per URL
No reference yet, or Accept current content enabledOK; the current content is stored as the reference
Content matches the referenceOK, Critical in reverse mode
Content differs from the referenceCritical until accepted, OK with a new reference in reverse mode
Status code outside 2xxCritical, error http-<status code>, for example http-503
Timeout exceededCritical, error timeout
Connection or certificate errorCritical, with the connection's error message
Entry is not an absolute http or https addressError invalid-url; the check counts as Failure

The following also applies to the check as a whole:

SituationResult
No URL enteredWarning, message no-urls
The stored references cannot be read or writtenFailure, message state-read-failed or state-write-failed
Unexpected errorFailure, message unexpected with details

The short result reads, for example, "Web content change — Critical: 1 of 3 changed". The detailed result text shows the number of URLs, lists every URL that is not OK with its error in parentheses, and shows "References established:" when references were stored on this run.

FAQ

The page returns slightly different content on every request, for example because of timestamps, random values or session tokens. Because the check compares byte for byte, that is a change every time. Monitor such pages with the HTTP Content Check and a text that must be present or absent.

The other URLs keep their reference. A new URL receives its first reference on the next run. If you remove a URL and add it again later, the check compares against its old reference again.

Yes. The reference applies per device, URL and mode. If two Web Content Change checks on the same device watch the same URL in the same mode, Accept current content in one of them also accepts the change for the other.