Database Observability

A module that brings PostgreSQL — instance availability and load, queries, locks, replication, and data access — into the same view as the rest of your infrastructure.

Database Observability

Database Observability extends SAF to PostgreSQL observability: instance availability and load, query performance, locks, replication and high-availability clusters, data growth, and data access control. The module links DBMS telemetry to Inventory assets, the Service Monitor Toolkit health model, and Incident Manager cards, so database health is analyzed in the same environment where the rest of your infrastructure is already operated.

Who it's for:

Database administrators and DBMS performance specialists
On-call teams and operations engineers
Business service owners and SREs
Information security teams, for data access auditing

key features

Single view of your database fleet

A consolidated picture of all monitored PostgreSQL instances, with one-click drill-down to the state of a specific instance:

availability of clusters and instances, the number of unavailable instances, and open incidents.
share of occupied connections, TPS, cache hits, and replication lag for each instance.
cluster, environment, criticality, node role, and fleet distribution by PostgreSQL version.
drill-down to initial instance diagnostics: uptime, transactions and rollbacks, sessions and wait types, I/O profile, and database sizes.

Incidents, SLAs, and AI analysis

The module delivers not only detection but also the process:

24 jobs register incidents: instance availability, connection saturation, query degradation, locks and deadlocks, replication and slots, XID consumption and database growth, and collection failures.
the card carries the incident context, initial diagnostics, operator recommendations, and an automatic link to the database asset.
severity is calculated by threshold, and the incident is grouped by object rather than by result row.
two SLA policies: response time and resolution time for capacity incidents.
AI explains slow query plans and summarizes load changes.

Database assets in Inventory and health in Service Monitor Toolkit

A breakdown of AI agent behavior and request structure down to the individual span:

requests, responses, tool calls, tokens, and latency across conversations.
drill-down to traces: span details, a request map, and errors.
pinpointing bottlenecks and root causes of failures at the tool-call level.