> ## Documentation Index
> Fetch the complete documentation index at: https://docs.upmetr.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Settings

> Configure how Upmetr monitors your infrastructure, polls metrics, and enforces security policies

The **Settings > General** page controls how Upmetr monitors your infrastructure. These are organization-level settings — each organization has its own independent configuration.

Navigate to **Settings > General** to access all options below.

<Note>
  You need **Admin** or **Superadmin** role to change settings.
</Note>

***

## Discovery Settings

Controls automatic scanning of your connected cloud accounts for new or changed resources.

| Setting                 | Description                                   | Default    | Options                                          |
| ----------------------- | --------------------------------------------- | ---------- | ------------------------------------------------ |
| **Automatic Discovery** | Toggle scheduled resource scanning on or off  | Off        | On / Off                                         |
| **Discovery Interval**  | How often Upmetr re-scans your cloud accounts | Every hour | Every hour, Every 6 hours, Every 12 hours, Daily |

When discovery is disabled, you can still trigger a manual scan from the **Cloud Accounts** page at any time.

<Tip>
  For accounts with frequently changing resources (auto-scaling groups, spot instances), use a shorter interval like **Every hour**. For stable environments, **Daily** reduces API calls.
</Tip>

***

## Uptime Monitoring

Configures HTTP endpoint and website monitoring. Uptime monitoring is always active — it cannot be disabled.

| Setting            | Description                       | Default    |
| ------------------ | --------------------------------- | ---------- |
| **Check Interval** | How often each monitor is checked | 60 seconds |

Each check performs up to 3 HTTP attempts (1 initial + 2 retries with 1-second delay) to avoid false positives from transient network issues. A monitor must pass a configurable **recovery threshold** (default: 2 consecutive successes) before auto-resolving an incident.

***

## CloudWatch Metrics

Controls polling of AWS CloudWatch for managed service metrics (RDS, ALB, CloudFront, etc.).

| Setting                | Description                                     | Default   | Options                  |
| ---------------------- | ----------------------------------------------- | --------- | ------------------------ |
| **CloudWatch Polling** | Global toggle for CloudWatch metrics collection | Off       | On / Off                 |
| **Polling Interval**   | How often metrics are fetched from CloudWatch   | 5 minutes | 5, 10, 15, or 30 minutes |

CloudWatch polling has two layers of control:

1. **Global toggle** — this setting page (applies to the entire organization)
2. **Per-account opt-in** — each cloud account must also have CloudWatch enabled on the Cloud Accounts page

Both must be enabled for metrics to flow.

<Note>
  CloudWatch API calls incur AWS costs (\~$0.01 per 1,000 metric data points). Estimated cost: ~$1–4/month per account depending on resource count. The IAM policy must include `cloudwatch:GetMetricData` and `cloudwatch:ListMetrics`.
</Note>

<Tip>
  GCP Monitoring and Azure Monitor polling can also be enabled per-account. These follow the same pattern as CloudWatch — a global toggle plus per-account opt-in.
</Tip>

***

## Session & Security

Authentication and access control policies for your organization.

| Setting                | Description                                                  | Default    |
| ---------------------- | ------------------------------------------------------------ | ---------- |
| **Session Timeout**    | Minutes of inactivity before a user is logged out            | 30 minutes |
| **MFA Enforcement**    | Require multi-factor authentication for all users            | Off        |
| **Max Login Attempts** | Failed login attempts before account lockout                 | 5          |
| **Lockout Duration**   | Minutes an account stays locked after exceeding max attempts | 15 minutes |

<Tip>
  Enabling **MFA Enforcement** is strongly recommended for production environments. Users will be prompted to set up TOTP on their next login.
</Tip>

***

## Data Retention

| Setting            | Description                                | Default |
| ------------------ | ------------------------------------------ | ------- |
| **Data Retention** | Number of days to keep metric and log data | 90 days |

Data older than the retention period is automatically purged. Adjust this based on your compliance requirements and storage needs.

***

## Settings vs Organization

Upmetr separates **monitoring behavior** from **organization identity**:

* **Settings** (this page) — controls how monitoring works: discovery intervals, polling, security policies, data retention.
* **Organization** (separate page at **Settings > Organization**) — controls branding and identity: organization name, custom logo, accent colors, and whitelabeling options.

See the [Organization guide](/guides/organization) for branding configuration.