Link Search Menu Expand Document

Revocation Monitor

This feature is available from certdog 1.18


The Revocation Monitor continually checks the OCSP responders and CRLs (Certificate Revocation Lists) that your certificates depend on, and alerts you when they fail

If an OCSP responder stops responding, or a CRL is not republished before it expires, relying parties can no longer check whether certificates have been revoked. Depending on the application, certificates may then be rejected (causing an outage) or accepted without a revocation check. The Revocation Monitor detects these problems early - usually before the CRL or OCSP response actually expires

It can monitor any OCSP responder or CRL: those provided by certdog’s own Local CAs and OCSP servers, Microsoft AD CS, public CAs and partner CAs

This replaces the stand-alone CRL OCSP Monitor Windows application


Overview

Each monitor checks one OCSP responder or one CRL location on a schedule (e.g. every minute):

  • OCSP monitors send an OCSP request for a certificate and check the response: that the responder answered in time, that the response is correctly signed by an authorised responder, that it is current and that the certificate has the expected status (e.g. Good)
  • CRL monitors download a CRL and check it: that it downloaded in time, that it is correctly signed by the CA, that it is current and has enough time left before it expires, and optionally that a given certificate is (or is not) on the list

Every check is recorded, so the history of each monitor can be viewed - for example when response times increased, or when a server behind a load balancer first started to fail

When a monitor starts failing, and again when it recovers, alerts can be sent by email, by webhook (e.g. to Microsoft Teams or Slack) and as certdog notifications

Both RSA, EC and post-quantum (ML-DSA, SLH-DSA) signatures are supported


Viewing Monitors

From the menu, select Monitoring and then choose Revocation Monitors

Revocation Monitors

The tiles at the top show how many monitors there are, how many are passing, failing and disabled, and how many checks have run in the last 24 hours

Each monitor is listed with its status, type, target URL, how often it runs, when it was last checked, its last response time and its response times over the last 24 hours. Use Search, Type and Status to filter the list

Click a monitor (anywhere on its row) to show its actions:

  • Run now - run the check straight away
  • History - view its history (see Monitoring History)
  • Edit - change its settings
  • Disable / Enable - stop or restart its checks. A disabled monitor is not checked and sends no alerts
  • Delete - delete the monitor and its history


Creating a Monitor

From the Revocation Monitors page, click Add Monitor

Only the certificate and issuer (OCSP) or the CRL location (CRL) are required. Everything else starts with sensible defaults


1. What to Monitor - OCSP

Add Monitor - OCSP

  • Type. Choose OCSP responder. The type cannot be changed after the monitor is saved
  • Name. A name for the monitor. This must be unique
  • Description. Optional
  • Certificate to check. The certificate the OCSP request is made for. Paste it (PEM), upload a file, or choose it from certdog (see below). It is best to use a certificate kept for monitoring, so that its status never changes unexpectedly
  • Issuer certificate. The CA certificate that issued the certificate to check
  • Expected result. The status the responder should return for the certificate: Good, Revoked, Unknown or Try later. For example, to check that revocations are being published, monitor a test certificate that has been revoked on purpose and expect Revoked
  • Responder URL. Leave this empty to use the OCSP URL in the certificate (its AIA extension). Enter a URL to check a specific responder, e.g. one node behind a load balancer


Choosing Certificates From Certdog

The Certificate to check and Issuer certificate each have a From certdog tab, to choose a certificate that is already in certdog instead of pasting it:

Choosing a certificate from certdog

Search by name or serial number and choose Active, Revoked or All certificates. Click Use to choose a certificate

When an issuer has already been chosen, the switch Only certificates issued by… limits the list to certificates that issuer really issued. The issuer list shows CA certificates only

When the certificate to check is chosen, certdog also:

  • Fills in its issuer, if certdog holds the issuer’s certificate (e.g. from a Local CA or a Certificate Issuer). If certdog does not hold it, you are asked to choose or paste it
  • Warns if a different issuer had already been chosen, and offers to use the correct one
  • Sets Expected result to Revoked if the certificate is revoked
  • Fills in the Name, if it is empty

The issuer filled in from certdog

When a monitor is saved, certdog checks that the issuer certificate really issued the certificate to check (by its signature). If not, the monitor is not saved and an error is shown


1. What to Monitor - CRL

Add Monitor - CRL

  • Type. Choose CRL distribution point
  • Name and Description. As above
  • CRL location. The URL of the CRL. http://, https:// and ldap:// URLs are supported. LDAP URLs must include the server name e.g. ldap://dc1.example.com/CN=Issuing CA,CN=CDP,...?certificateRevocationList?base
  • CRL issuer. The CA certificate that signs the CRL. This can also be chosen From certdog
  • Verify the CRL signature. Checks that the CRL is signed by this issuer and names it as the issuer. This is recommended
  • Check a certificate is on the CRL. Optional. Choose the certificate to look for and whether it should be Revoked (listed) or Good (not listed). For example, a test certificate revoked on purpose proves the CRL is current


2. Schedule

Schedule

  • Enabled. When unchecked, the monitor is saved but not checked
  • Check every. How often the check runs. The minimum is 30 seconds (this can be changed, see Settings)
  • Timeout. How long to wait for a response. Leave empty for the default (30 seconds for OCSP, 60 for CRLs)
  • Skip weekends. No checks are run on Saturday or Sunday (server time)
  • Use proxy. The check is made through the proxy set in the Revocation Monitor Settings. Use this for responders and CRLs on the internet when certdog cannot reach them directly


3. Checks

The checks that are always made are described above. Further checks can be added here

OCSP

OCSP checks

  • Response time. Fail if the response takes longer than this
  • Response lifetime. Fail if the time between the response’s thisUpdate and nextUpdate is more (or less) than this
  • Responder certificate expiry. Fail if the responder’s signing certificate expires within this many days
  • Send a nonce. Include a nonce in the request. The response must echo it back. Note that many responders (and all that pre-produce responses) do not support nonces
  • Require OCSP Signing usage. For responders that sign with a delegated certificate, the certificate must have the OCSP Signing extended key usage (RFC 6960). This is recommended

Under Advanced: request format, clock skew, trusted responders:

  • HTTP method. POST (the default) or GET
  • CertID hash. The hash used to identify the certificate in the request. SHA-1 is the most compatible
  • Allowed clock skew. Allow for clock differences between certdog and the responder, in seconds
  • Trusted responder certificates. Only needed when the responder signs with a certificate that the issuing CA did not issue

CRL

CRL checks

  • Time left before nextUpdate. Fail if the CRL has less (or more) than this time left before its nextUpdate. Check Measure to Next CRL Publish to measure to the CRL’s Next CRL Publish time instead, when present (as published by Microsoft AD CS). This warns that a CRL has not been republished well before it expires
  • Download time. Fail if the download takes longer than this
  • CRL size. Fail if the CRL is larger (or smaller) than this. An unexpectedly small CRL can indicate a problem with publishing
  • Allowed clock skew. Under Advanced: clock skew


4. Alerts

Alerts

  • Alert after. An alert is sent once this many checks in a row have failed. Higher values ignore one-off blips, e.g. a single node behind a load balancer failing once
  • Alert on every failed check, not only the first. By default one alert is sent when the monitor starts failing. Check this to be alerted on every failed check
  • Alert on recovery. Send an alert when a monitor that had alerted starts passing again
  • Email recipients. The email addresses to send alerts to. Certdog’s Email Settings must be configured
  • Webhooks. URLs to send a JSON message to, e.g. a Microsoft Teams or Slack incoming webhook. See MS Teams Notifications for how to create a Teams webhook
  • Notifications. Send a certdog notification to Everyone, to chosen Teams or to chosen Users. See Alerts below

Under Customise email text and webhook payload, the email subjects and bodies and the webhook payload can be changed. Leave these empty for the defaults. The following tokens can be used, and are replaced with the values from the check (click a token to copy it):

Token Value
[monitorname] The monitor’s name
[url] The URL checked
[failures] Why the check failed
[checkedat] When the check ran
[expected] The expected certificate status
[observed] The status returned
[responsetimems] The response time in milliseconds
[consecutivefailures] How many checks in a row have failed
[runby] The certdog server that ran the check
  • Webhook header name / value. An optional HTTP header sent with webhooks, e.g. for authentication
  • Webhook payload. The JSON sent to the webhooks. If empty, a default JSON payload with the details of the check is sent
  • Send the payload as a JSON string. Sends the payload as a single JSON string value (for services that expect this)


5. History

History settings

  • Keep results for. How long this monitor’s results are kept. System default uses the server setting (90 days by default, see Settings). Choose Custom to keep this monitor’s results for a different number of days


Testing and Saving

Click Test to run the check once, with the settings entered, without saving anything. The result is displayed, including the full transcript of the check:

A test result

Click Save monitor to save it. The first check runs straight away


Monitors for Local CAs

Monitors can be created for certdog’s own Local CAs, with the details filled in from the CA


When Creating a CA

When creating a new Local CA with Generate CRLs checked, the option Monitor this CRL is shown (checked by default):

Monitor this CRL

When the CA is created, a CRL monitor is added for each of its CRL distribution points (HTTP and LDAP). The CA’s certificate is used as the issuer, the signature is checked, and the remaining lifetime alert is set from the CA’s CRL lifetime and generation settings. The first check runs about two minutes after the CA is created, by which time its first CRL has been published

This option is not available for an intermediate CA whose certificate is requested from an external CA (CSR). Add the monitor from the CA’s page (below) once its certificate has been accepted


For an Existing CA

From Local CAs, choose CA Configuration, select the CA and choose Edit. The Revocation monitoring section shows the monitors that already watch the CA’s CRL and OCSP URLs:

Revocation monitoring for a Local CA

  • Create CRL monitor. Adds a CRL monitor for each of the CA’s CRL distribution points that is not already monitored
  • Create OCSP monitor. Opens a new OCSP monitor with the name, issuer and OCSP URL filled in from the CA. Choose the certificate to check, from the list of certificates the CA has issued, and click Save monitor. This is only available once the CA has issued at least one certificate, as an OCSP check needs a real certificate from the CA


Editing, Disabling and Deleting a Monitor

Editing. From the Revocation Monitors page, select the monitor and choose Edit (or click Edit on its history page). Change the settings and click Save monitor. The monitor is checked straight away with the new settings

Disabling. Select the monitor and choose Disable. It is not checked and sends no alerts until it is enabled again. Its history is kept. Alternatively, uncheck Enabled when editing the monitor

Deleting. Select the monitor and choose Delete, then click Yes to confirm. The monitor and all of its history are deleted. Any notification it raised is removed


Monitoring History

From the Revocation Monitors page, click the monitor and choose History:

Monitor history

  • Period. Show the last 24 hours, 7 days, 30 days, 90 days, or a Custom range
  • Group by. How finely the charts group results, e.g. per minute or per hour. Auto chooses a suitable size for the period

The tiles show, for the period:

  • Availability. The percentage of checks that passed
  • Checks. How many checks ran
  • Failed checks. How many checks failed, and how many times the monitor started failing
  • Average response and Slowest response
  • Signer cert expires (OCSP) or Next update (CRL)

The charts show:

  • Response time. The average response time (line), with the fastest to slowest response shaded. Use this to see when response times changed
  • Checks failed. The share of checks that failed in each period. 100% means the endpoint was down; a lower bar means only some checks failed, e.g. 50% when one of two servers behind a load balancer is failing

Pass/fail changes lists each time the monitor started failing and when it passed again, with the reason

Results lists the individual checks. Choose All, Failures or Changes only, and use the page controls to move through them. Click Details to view a check

On the history page you can also Run now, Edit the monitor, or Clear history (delete all of its results)


Check Details

Each check’s details show why it failed, the details of the check and of the response, and the full transcript of the check:

Check result

For failed checks the raw OCSP response is kept, and can be copied (base64) or downloaded (.der) to examine with other tools e.g. openssl ocsp -respin response.der -resp_text

By default the transcript and raw response are kept for failures and for checks where the monitor changed between passing and failing. The pass/fail result, response time, size and validity of every check are always kept. See Detail kept in Settings


Alerts

Alerts are sent when a monitor starts failing (once the Alert after number of checks in a row have failed) and, if Alert on recovery is set, when it passes again. With Alert on every failed check set, an alert is sent for each failed check

All alerts are also written to the certdog log


Email

Emails are sent to each of the Email recipients. The default email includes the monitor’s name and type, how many times it has failed in a row, the URL, when and by which server it was checked, the expected and received status, and why it failed. The recovery email includes the response time

The email text can be changed with the tokens listed above. Email is sent using certdog’s Email Settings


Webhooks

A JSON message is posted to each webhook URL. By default this contains:

{
  "event": "failure",
  "monitor": "Partner CA OCSP",
  "monitorId": "6740...",
  "type": "OCSP",
  "url": "http://ocsp.partner.example.com",
  "passed": false,
  "checkedAt": "2026-11-21T12:03:10Z",
  "consecutiveFailures": 3,
  "failures": "The OCSP response is out of date: nextUpdate (2026-11-20T10:58:00Z) is in the past"
}

The event is failure or recovery. Use a custom Webhook payload for services that need a particular format, such as Microsoft Teams or Slack


Notifications

A certdog notification can be sent to Everyone, to chosen Teams, or to chosen Users. It appears in the notification bell at the top of the certdog console

Each monitor raises one notification, which follows its state:

  • When the monitor starts failing, a Critical notification is raised, e.g. Revocation monitor failing: Partner CA OCSP, with the reason
  • If it keeps failing (with Alert on every failed check), the same notification is updated rather than adding more. The number of times it has occurred is shown
  • When it recovers, the same notification is updated to Info, e.g. Revocation monitor recovered: Partner CA OCSP

Each update shows as unread again. The notification is removed when the monitor is deleted or when its notifications are turned off

image-20261121135823186


The Dashboard

Administrators see the Revocation Monitor on the certdog Dashboard: a summary of passing and failing monitors, and the status of each monitor over the last 24 hours (one square per hour), with what expires next and recent changes:

Revocation monitoring on the Dashboard


Settings

From the Revocation Monitors page, click Settings

Revocation Monitor Settings

Proxy. The proxy used by monitors that have Use proxy set, for servers that cannot reach the internet directly. Monitors for internal CAs normally go direct. This applies to every certdog server

  • Proxy host and Port. Leave the host empty for no proxy
  • Username and Password. Optional. The password is stored encrypted and is never displayed again. Click Change to replace it, or check Remove password

Basic proxy authentication over HTTPS is disabled by the Java runtime by default. If an https:// endpoint fails with an authenticated proxy, clear jdk.http.auth.tunneling.disabledSchemes in the Java options

Server settings are set in certdog’s application.properties file on each server, and are displayed here for reference. All are optional. The defaults are:

Setting Default Description
revmon.thisserver.run true Set to false to stop this server running checks. With several certdog servers, checks are shared between those that have this set to true. Each check runs once, on one server
revmon.threads 4 The maximum number of checks run in parallel on this server
revmon.mininterval.seconds 30 The shortest interval allowed between checks of one monitor
revmon.results.retentiondays 90 Results older than this many days are deleted. A monitor can override this
revmon.results.purge.cron 0 25 3 * * ? When old results are deleted (3:25am every day)
revmon.results.detail FAILURES How much detail is stored with each result: ALL, FAILURES (failures and changes between passing and failing) or NONE. The pass/fail result, response time, size and validity are always stored
revmon.results.maxtoreturn 5000 The maximum number of results returned by one history request
revmon.allowfileurls false Allow file:// CRL locations (reads files on the certdog server)
revmon.proxy.host / .port / .username / .password   A proxy to use when none is set on the Settings page

Restart certdog after changing these settings


REST API

Monitors can also be managed with the REST API. See the API documentation (Swagger) on your certdog server for the details:

Method Path Purpose
GET /admin/revmon List the monitors
POST /admin/revmon/add Add a monitor
GET /admin/revmon/{id} Get a monitor
PATCH /admin/revmon/{id}/update Update a monitor
DELETE /admin/revmon/{id} Delete a monitor and its history
POST /admin/revmon/test Run a check without saving it
POST /admin/revmon/{id}/run Run a monitor now
GET /admin/revmon/{id}/results Get a monitor’s results
GET /admin/revmon/{id}/stats Get a monitor’s statistics, grouped by time
GET /admin/revmon/summary The status of all monitors (as on the Dashboard)
GET / PATCH /admin/revmon/settings Get or update the proxy settings
POST /admin/revmon/certificates/search Search certdog’s certificates to monitor
GET /admin/revmon/localca/{caId}/template The details needed to monitor a Local CA
POST /admin/revmon/localca/{caId}/crl-monitors Add CRL monitors for a Local CA