Monitoring & Observability Integrations

The Monitoring tab of the Integrations page connects Breeze to the observability and on-call tools your team already runs. From a single screen you can publish Breeze metrics to a Prometheus scrape endpoint, deep-link to Grafana dashboards, route critical alerts to PagerDuty and OpsGenie, forward monitoring events to any number of outbound webhooks, and choose exactly which metrics are exposed to external monitors.

Open it from Integrations → Monitoring (/integrations#monitoring).

Not the internal Monitoring stack

This page is about sending Breeze’s telemetry and alerts out to your third-party tools. It is not the self-hosted observability stack that monitors the Breeze platform itself. If you are looking to stand up Prometheus, Grafana, Loki, and Alertmanager to watch the Breeze API/agents, see the Observability Stack instead — that’s a different thing entirely.

Per-organization settings

Monitoring integrations are configured per organization. Select a specific organization first — while the All orgs view is active there is no single org to load or save against, and the tab shows a prompt asking you to switch to one organization to view or edit its settings.

Each provider block has its own enabled toggle, a Test connection button that reports a live status badge (Not tested / Testing / Active / Failed), and a per-provider result message. Nothing is persisted until you click Save settings in the header — an Unsaved changes indicator appears while you have pending edits.

---

Prometheus scrape endpoint

Publish Breeze metrics so a Prometheus server can scrape them on a schedule you control.

Field	Description
Endpoint URL	The base URL where the metrics are published (e.g. https://prometheus.example.com).
Metrics path	The HTTP path the metrics are served on (default /metrics).
Scrape interval	How often Prometheus should scrape, as a duration string (e.g. 30s).
Scrape timeout	How long a single scrape may take before it is abandoned (e.g. 10s).

1. In Integrations → Monitoring, find the Prometheus card and turn its toggle to Enabled.

2. Set the Endpoint URL and Metrics path for where metrics should be published, then set the Scrape interval and Scrape timeout to match your Prometheus scrape config.

3. Choose which metrics are exposed in the Metric export panel — only selected metrics are surfaced to external monitors.

4. Click Test connection to verify reachability. The status badge turns Active on success or Failed with a message on error.

5. Click Save settings in the page header to persist the configuration.

Tip

The Scrape interval and Scrape timeout are plain duration strings — use the same format your Prometheus scrape_configs expect (15s, 30s, 1m, and so on).

---

Grafana

Link your Grafana instance so Breeze can deep-link into the right dashboards.

Field	Description
Dashboard URL	The base URL of your Grafana instance (e.g. https://grafana.example.com).
API key	A Grafana API key/token used to authenticate. Stored as a secret field (masked input).
Organization ID	The numeric Grafana org ID that owns the dashboards (Grafana’s own org identifier, e.g. 1).

Toggle the Grafana card Enabled, fill in the three fields, use Test connection to confirm the instance is reachable, and Save settings.

Note

The Organization ID here is Grafana’s internal org number — it is unrelated to your Breeze organization. Leave it at 1 unless your Grafana setup uses multiple orgs.

---

PagerDuty

Trigger on-call workflows from critical Breeze alerts.

Field	Description
Integration key	The PagerDuty Events API integration (routing) key. Stored as a secret field (masked input).
Service ID	The PagerDuty service the alerts should be attached to.

OpsGenie

Route incident responses to the right OpsGenie team.

Field	Description
API key	The OpsGenie API key. Stored as a secret field (masked input).
Team	The OpsGenie team that should receive the routed incidents.

Wiring up PagerDuty or OpsGenie

1. In your PagerDuty or OpsGenie account, create an integration and copy its key — a PagerDuty Events API integration key, or an OpsGenie API key.

2. In Integrations → Monitoring, toggle the PagerDuty or OpsGenie card to Enabled (both start disabled).

3. Paste the key into Integration key (PagerDuty) or API key (OpsGenie), then add the Service ID (PagerDuty) or Team (OpsGenie) that should own the incidents.

4. Click Test connection and confirm the badge reads Active.

5. Click Save settings to persist.

Tip

The Test connection button is disabled while a provider is toggled off — enable the provider first, then test.

---

Outbound webhooks

Forward monitoring events to any number of external HTTP endpoints — a Datadog relay, a CloudWatch bridge, or your own ingestion service. Each endpoint has:

Field	Description
Name	A label for the endpoint (e.g. “Datadog relay”).
Endpoint URL	The destination URL events are forwarded to.
Active	A per-endpoint checkbox controlling whether this endpoint receives events.

Manage endpoints from the Custom monitoring webhooks card:

• Add an endpoint — enter a Name and Endpoint URL in the row at the bottom of the card and click Add endpoint.
• Test one endpoint — click Test on that endpoint’s row to send it a probe; its own status badge reports the result.
• Test all — the Test all button in the card header probes the whole set at once.
• Remove an endpoint — click Remove on its row.
• Enable/disable the whole card — the card’s Enabled toggle gates all outbound webhook delivery for the org.

Note

These outbound monitoring webhooks are the lightweight per-org forwarders configured on this tab. They are separate from Breeze’s full-featured, signed, retrying event webhooks, which offer HMAC signatures, delivery history, and automatic retries. Use event webhooks when you need guaranteed, verifiable delivery of specific event types.

---

Metric selection

The Metric export panel controls which metric families are exposed to external monitors (including the Prometheus scrape endpoint above). Toggle the panel on with its Enabled switch, then check the metrics you want to publish. A counter at the bottom shows how many are selected.

Metric	What it covers
Device health	Heartbeat, uptime, and connectivity
Performance	CPU, memory, and disk utilization
Patch compliance	Patch install rates and drift
Backup status	Job success rates and restore points
Security posture	Risk scores and posture checks
Automation runs	Run counts, duration, and failures
Alerts & incidents	Open alerts and escalations

Note

Only the metrics you check are surfaced to external monitors. Unchecking a metric family stops it from being exported without affecting how Breeze collects or displays it internally.

---

Saving and testing

• Save settings (header button) persists every provider block, the webhook endpoint list, and the metric selection together. An Unsaved changes note appears while edits are pending, and a confirmation banner shows once the save succeeds.
• Test connection / Test / Test all send a live probe using the values currently in the form (including unsaved edits) and report success or a failure message per provider or per webhook endpoint.

---

Related

Alerts & Rules

Define the alert rules and severities that feed PagerDuty and OpsGenie routing. See Alerts & Rules.

Webhooks

Breeze’s full signed, retrying event-webhook system with delivery history. See Webhooks.

Observability Stack

The self-hosted Prometheus/Grafana/Loki/Alertmanager stack that monitors the Breeze platform itself — a different thing from this tab. See Observability Stack.