# Create a monitor when you deploy a new service

When a new service goes live, add a monitor for it in the same pipeline, so nothing reaches production without
being watched. Running the step again does not create a second monitor.

**Runnable code:** the code on this page, ready to clone and run, is in [`infranest-examples/examples/monitor-per-deploy`](https://github.com/InfraNest-Infrastructure-Organized/infranest-examples/tree/main/examples/monitor-per-deploy). CI runs it against the [sandbox](../sandbox.md).

## The token

Make a token with **Monitoring → Write**.

## Create the monitor

```sh
curl -fsS -X POST https://api.infranest.io/api/organizations/42/monitoring-checks \
  -H "Authorization: Bearer $INFRANEST_TOKEN" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: monitor-api.example.com" \
  -d '{
    "name": "API health",
    "type": "http",
    "target": "https://api.example.com/health",
    "interval_seconds": 60,
    "ssl_check_enabled": true,
    "ssl_expiry_days": 14
  }'
```

- **`type`** is one of `http`, `port`, `ssl`, `dns`, `keyword`, `redirect`, `ping` or `api`.
- **`interval_seconds`** is one of 30, 60, 120, 300, 600, 900, 1800 or 3600. Your plan sets the shortest you
  can use.
- **The `Idempotency-Key`** is tied to the target, so a pipeline that runs twice within 24 hours creates the
  monitor once. For a check that should exist forever, first look for it:
  `GET /organizations/42/monitoring-checks`, and only create it when it is not there.

The response is the new monitor, including its `id`. Keep it if you want to pause the monitor around later
deploys.

## Pause it around a deploy

```sh
curl -fsS -X POST https://api.infranest.io/api/organizations/42/monitoring-checks/1201/pause \
  -H "Authorization: Bearer $INFRANEST_TOKEN" -H "Accept: application/json"

# … deploy …

curl -fsS -X POST https://api.infranest.io/api/organizations/42/monitoring-checks/1201/resume \
  -H "Authorization: Bearer $INFRANEST_TOKEN" -H "Accept: application/json"
```

For a deploy that also touches servers or several monitors, a
[maintenance window](maintenance-window-from-a-pipeline.md) is usually the better fit.