Administer
Platform health
Platform health is MonitOps monitoring itself. Its own components are watched with the same detection that watches your services, and failures are shown in the console as product states instead of being hidden.
How it works#
What is watched#
| Condition | Where it appears |
|---|---|
| Ingestion gaps | On the affected source |
| Stale sources | On the source, as acquisition freshness |
| Integration failures | On the integration or destination |
| Delivery backlogs | On the destination's queue |
| Degraded capabilities | As a degraded state on the capability |
Receivers for MonitOps' own nodes and databases supply host, database and storage signals, with catalogue entries, seeded alarm definitions, one infrastructure dashboard and a time-synchronisation check. Each service has liveness and readiness probes.
Circuit breakers#
Calls to external systems are protected by circuit breakers: fault-management delivery, ticketing delivery, the language model endpoint and calls to providers. A failing dependency is shown on the integration (see Destinations).
Degraded capabilities keep their contracts#
A capability that cannot run at full strength does not remove what it promised. It reports that it is degraded and delivers what it still can.
| If this is unavailable | What continues |
|---|---|
| The investigator | The computed cause ranking appears, marked as such (see Cause assessment) |
| The language model endpoint | The copilot says so; detection, correlation and the computed ranking continue |
| A destination | Its queue grows; other destinations and ingestion continue (see Destinations) |
Alerts from the platform#
Some platform conditions are alarm definitions like any other. For example, persistent log-forwarding failures raise a major alert that follows the normal path into incidents (see Log forwarding). A source that stops sending is raised by source silence (see Source health).
What you see in the console#
Source health, destination backlogs and degraded states appear in the console alongside your own services. There is no separate operator-only page for them.
Limits and defaults#
- Source silence is five times the expected interval, by default (see Source health).
- Missing data is never treated as healthy.
Failure states#
| State | What the console shows |
|---|---|
| Source silent | Silence state and acquisition freshness on the source |
| Destination failing | Backlog, errors and freshness on the destination |
| Investigator unavailable | Computed ranking, marked as computed |
| Language model unavailable | The copilot reports that it is unavailable |
Related#
Last updated October 2026