The monitor that knows which node is primary.
SQL Server on VMs with Always On, Azure SQL Database, Managed Instance, PostgreSQL Flexible Server and MariaDB, in one entity tree with one set of thresholds. Waits, queries, replicas, backups, linked servers and the reports that depend on them.
The things a DBA checks by hand at 3 a.m.
Primary and secondary carry their own metrics: log send and redo queues, synchronization state, failover history, listener reachability from where the apps actually live. A planned failover is an event, not an outage.
Select a window on the timeline and see the wait categories, the top queries and the open checks for exactly that span. Instance, wait and query data without the sampling gaps.
The collector runs the same catalog query your reports run, through the same linked-login mapping, on every replica. DB2 on an iSeries, Oracle, another SQL Server: if a report depends on it, Excubius checks it.
The housekeeping that drifts between tickets: last full and log backup, Query Store capacity, compatibility level against your policy, and the build each node runs, with Advisories flagging when they diverge.
Azure SQL Database and Managed Instance come in through Azure Monitor and the resource graph; PostgreSQL Flexible Server through its metrics and a read-only role. Same cards, same thresholds, same inbox.
The collector's login needs VIEW SERVER STATE and nothing more; every query it can run is published in the catalog. Your security team can read the list before you install anything.
101 checks in 13 modules, every one with a vendor baseline.
Each check ships as YAML: the query, how it is evaluated, and the threshold Microsoft, PostgreSQL or MariaDB documentation recommends. Thresholds inherit down your entity tree and can be overridden per resource.
| Module | Example checks | Source |
|---|---|---|
| SQL Server instance | CPU, PLE, waits by category, blocking, tempdb, agent jobs failed | collector · New Relic |
| Availability groups | Sync state, log send and redo queue, failover, listener DNS, replica build drift | collector |
| Databases | Backups, compatibility level, Query Store size, file growth, log reuse wait | collector |
| Linked servers and SSRS | Linked server reachable, catalog query time, report render, subscription failures | collector · Synthetics |
| Azure SQL and Managed Instance | DTU and vCore usage, storage, deadlocks, geo-replication lag, failover group state | Azure Monitor |
| PostgreSQL and MariaDB | Connections, replication lag, bloat, long transactions, slow queries, binlog or WAL growth | collector · Azure Monitor |
| Patch and lifecycle | New CU or GDR available, build behind policy, edition and end of support | Advisories |
What this release supports
Point it at your cluster. Watch the first alert arrive.
Databases is included in Free (up to 10 hosts) and in Pro. The trial covers every host and every product for 30 days.