How Does Certificate Monitoring Work?
Generator Labs SSL certificate monitoring works by regularly connecting to your monitored services, retrieving their TLS/SSL certificates, validating the certificate chains, and alerting you of any issues or approaching expirations.
The Monitoring Process
When a check runs, the system establishes a connection to the monitored service using the specified protocol (HTTPS, SMTPS, LDAPS, etc.). For protocols using STARTTLS, the system first connects to the unencrypted port, then issues the STARTTLS command to upgrade the connection. Once the TLS handshake completes, the system reads the certificate chain the server sends: its certificate and any intermediate certificates. The root comes from the trust store (or your private CAs); missing intermediates aren't fetched, so a server that doesn't send them is flagged.
The retrieved certificates are then validated: expiration dates are checked against configured alert thresholds, the certificate chain is verified as complete and properly linked, the hostname is matched against the certificate's Common Name or Subject Alternative Names, the signing CA is confirmed as trusted (public or configured private CAs), revocation is checked against the CA's certificate revocation list (CRL), and cryptographic algorithms are checked for weak or deprecated methods. The domain's DNS CAA records are checked too, and when the hostname resolves to several addresses, the certificates they serve are compared.
Findings
Each problem a check finds is a finding, with a key (such as CERT_HST02), a severity (notice, warning, error or critical) and a page explaining how to fix it. When a connection fails, only the earliest failure on each address is a finding, and the checks that need a certificate are shown as Not checked. The Findings page lists the findings across all your monitors.
Every check also notes what the server says about itself, to identify the software it runs; see what our checks look like in your logs. Changes to the certificate, its issues and the server's software are recorded in the monitor's History.
Alert Generation
When a check finds a new warning or error, or one clears, the system sends an alert to all contacts in the contact groups associated with the monitor. Alerts also go out when the certificate reaches one of the profile's expiration thresholds, when the certificate changes, and when flapping starts or ends. Notices (suggestions) never alert, and each group of findings can be turned off on the monitor's profile. The notification includes the issue, the affected monitor, and the certificate details.
An alert is sent when something changes, not on every check while a finding lasts; when the findings that alerted have cleared, you get an all-clear. A finding you've dismissed on a monitor doesn't alert. For expiration, you're alerted at each configured threshold (for example, at 60, 30, 15 and 5 days before expiration), and at 0 once the certificate has expired.
Monitoring Through Agents
When monitoring agents are configured and assigned to a profile, checks are routed through the agent instead of connecting directly from Generator Labs infrastructure. The agent receives the check request, connects to the target service from within your network, retrieves the certificates, and reports the results back to the platform. This enables monitoring of internal services while maintaining centralized alerting and reporting. See installing monitoring agents for deployment instructions.
Connection Sources
By default, monitoring checks originate from Generator Labs infrastructure over the public internet. When using DNS hostnames, the system performs A and AAAA record lookups and monitors all returned IP addresses, making it ideal for load-balanced services. For services on private networks, deploy monitoring agents to enable access without exposing services to the public internet.