> For the complete documentation index, see [llms.txt](https://docs.upriverdata.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.upriverdata.com/product-usage/on-call.md).

# On-call

Triage alerts and incidents from across your data stack in one place.

On-call collects every alert that reaches Upriver — alerts from the [Slack](https://docs.upriverdata.com/integrations/slack.md) and [Microsoft Teams](https://docs.upriverdata.com/integrations/microsoft-teams.md) channels you activate, alerts forwarded by tools like [Monte Carlo](https://docs.upriverdata.com/integrations/monte-carlo.md), Airflow (Astronomer or Cloud Composer) and Fivetran, and alerts raised by Upriver's own monitors — and investigates each one against your actual data.

Related alerts are grouped into **incidents**. Each incident gets a summary, a root cause, an impact assessment, and action items, so you can triage from one page instead of chasing alerts across tools.

Open On-call from the side navigation.

<figure><img src="https://875415170-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FLE0zsJmUpeLjYw085cnO%2Fuploads%2Fgit-blob-e9b27e720419aad0af36af3e79e03a16ab89472e%2Fon-call-overview.png?alt=media" alt=""><figcaption><p>The On-call page: summary cards above the incident list</p></figcaption></figure>

## The page layout

The top of the page shows four summary cards for the current filter:

* **Unresolved incidents** — how many open incidents still need triage.
* **New incidents** — incidents that arrived in the last 24 hours, with a 7-day trend.
* **Alerts by source** — which tools the incoming alerts came from.
* **Incident statuses** — a breakdown of the filtered incidents by status.

Below the cards, two tabs switch between views:

* **Incidents** — the grouped incidents list: see [Incidents](https://docs.upriverdata.com/product-usage/on-call/incidents.md).
* **Alerts** — the raw alert stream: see [Alerts](https://docs.upriverdata.com/product-usage/on-call/alerts.md).

A date range picker at the right of the filter bar limits both tabs to a window such as the last 2 days, and shows how many records match.

## How alerts become incidents

Every alert that arrives is analyzed by Upriver. When an alert matches an existing open incident — same pipeline, same failing job, same upstream cause — it is attached to that incident instead of creating a new one. Alerts that don't match anything become a new incident, which Upriver investigates and classifies with a status and a severity (see [Status and Severity](https://docs.upriverdata.com/product-usage/on-call/status-and-severity.md)).

Clicking an incident opens the [incident page](https://docs.upriverdata.com/product-usage/on-call/incident-page.md): a full investigation with the root cause, supporting evidence, action items and an activity timeline.
