Core concepts#
A short glossary. You do not need to read this first, but it explains terms the rest of the docs assume.
Organisation and tenant#
An organisation is your workspace. All your dashboards, monitors, alerts and connected systems belong to it. Each organisation maps to an isolated tenant, and that isolation is applied where the data is stored, not only in the application. One organisation can never query another's data.
Larger customers often run several organisations, for example one per business unit, each with its own settings, retention and access.
Signals#
| Signal | What it answers |
|---|---|
| Metrics | How much, how many, how fast, over time |
| Logs | What exactly happened, in detail, at one moment |
| Traces | Where the time went across services for one request |
| Profiles | Which functions inside a process consumed CPU or memory |
Data sources#
A data source is a system ObserveIQ can query. It might be built in, or a system you already run such as your own Prometheus or a SQL database. Once connected, a data source is available on dashboards, in Explore and in alert rules, with no further configuration.
Monitors#
A monitor is an active check that runs on a schedule from outside your system: an HTTP request, a TLS certificate check, a DNS lookup, a TCP connection or a ping. Monitors produce the availability record used by service level objectives and status pages.
Alert rules, contact points and policies#
An alert rule defines a condition. A contact point is somewhere a notification can go, such as an email address or a PagerDuty service. A notification policy decides which alerts go to which contact points, based on severity and category.
This separation is the point: routing is configured once, centrally, rather than repeated on every rule. Adding a new destination does not mean editing every alert you have.
Service#
A service is a named application that emits telemetry, taken from the service.name attribute in OpenTelemetry. Getting this name right matters, because it is the key that joins traces, logs, metrics and profiles for the same application. See Service view.