Skip to main content
Version: 0.4 · current

Troubleshoot availability group monitoring

TroubleshootingReviewed October 11, 2026
Applies to:Excubius 0.4Roles: Entity admin, Editor, Operator
PlansFree · Pro · EnterpriseDeploymentSelf-hosted

Each section starts with what you see, then explains the likely cause and how to fix it. Start with the error kind shown on the check: it narrows the cause.

Error kind on the checkMeans
authLogin failed, or the login cannot open the database it needs
unreachableNetwork error, connection refused or dropped
timeoutLogin or connection timeout
querySQL Server answered but the statement failed, usually a missing permission
configSomething to fix in Excubius: no host, no or wrong credential, unreadable secret

Checks fail with auth after a failover​

Symptom. Everything was healthy; after a failover, checks on the group fail with auth.

Cause. The monitoring login was created on the new primary with a different SID, or not created at all, so it no longer maps to its users in msdb and in the availability databases.

Resolution. Recreate the login on that replica with the SID of the first replica, then repeat the grants and the msdb part. See Same login on every replica.

The listener check is critical but applications work​

Symptom. Listener IP online is Critical, while applications connect normally.

Cause. In a multi-subnet group, only one listener IP is online at a time. If the Excubius server cannot reach the IP of the subnet where the primary now runs, the check cannot see it.

Resolution. Allow the Excubius server to reach every listener IP on the SQL port, not only the one that was online when you set it up.

Per-database checks skip the availability databases​

Symptom. Instance checks work, but per-database checks report nothing for the databases in the group.

Cause. Excubius connects with ApplicationIntent=ReadOnly. A primary configured with PRIMARY_ROLE (ALLOW_CONNECTIONS = READ_WRITE) refuses read-intent connections to its availability databases.

Resolution. Set the credential attribute application_intent: ReadWrite. It only changes routing; the login still has no write permission.

A resource shows a single "Cannot reach" alert​

Symptom. One critical alert Cannot reach appears on a replica, and its other alerts list it as their root cause.

Cause. By default, three consecutive checks failed with unreachable, auth or timeout. Excubius opens one alert instead of many and holds the notifications of the dependent alerts.

Resolution. Fix the connection: test it from the resource page with Test connection. The alert resolves by itself as soon as the checks succeed again.

Backup preference reports a mismatch you cannot see​

Symptom. Backup preference honoured warns, but the backup jobs look right.

Cause. Through the listener, the check reads msdb on the primary. It warns when the primary took full backups of the group's databases in the last 26 hours although it is not their preferred backup replica. Backups taken on other replicas are recorded in their own msdb and are not visible from there.

Resolution. Check which replica runs the full backup job and compare with the group's automated backup preference. Run backups where the preference says, or change the preference.

Still stuck​

Signed-in users can read Known issues and limitations. When you contact support, include the check id, the error kind and the time it started.