> ## 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.
# Resources & Compute Management
> Discover, inventory, and manage cloud resources across AWS, GCP, Azure, and DigitalOcean
Upmetr automatically discovers and inventories cloud resources across all connected providers. The **Resources** page gives you a unified view of every resource in your infrastructure — compute instances, databases, DNS zones, storage buckets, serverless functions, and more — regardless of which cloud they live in.
## Resource Discovery
When you [add a cloud account](/guides/cloud-accounts), Upmetr immediately scans it for supported resource types. After the initial scan, discovery runs automatically every **15 minutes** to keep your inventory up to date.
Each discovered resource includes:
* **Name and ID** — The resource's name and provider-specific identifier
* **Type** — The resource category (e.g., EC2 Instance, Cloud SQL, Droplet)
* **Status** — Current state (Running, Stopped, etc.)
* **Region** — The cloud region where the resource is deployed
* **Cloud Account** — Which connected account owns the resource
Resource discovery requires the **cloud\_resources** module. If your plan does not include this module, the Resources page will show an upgrade prompt.
## Supported Resource Types
Upmetr discovers 27 resource types across four cloud providers:
### AWS
| Type | Category |
| ----------------------- | ----------- |
| EC2 Instance | Compute |
| RDS Database | Database |
| Route 53 Zone | DNS |
| CloudFront Distribution | CDN |
| Load Balancer (ALB) | Network |
| ACM Certificate | Certificate |
| S3 Bucket | Storage |
| Lambda Function | Serverless |
| WAF Web ACL | Security |
### Google Cloud
| Type | Category |
| -------------------- | ---------- |
| Compute Engine VM | Compute |
| Cloud SQL Instance | Database |
| GKE Cluster | Kubernetes |
| Cloud DNS Zone | DNS |
| Cloud Storage Bucket | Storage |
| Cloud Function | Serverless |
### Azure
| Type | Category |
| ------------------ | ---------- |
| Azure VM | Compute |
| Azure SQL Database | Database |
| AKS Cluster | Kubernetes |
| Azure DNS Zone | DNS |
| Blob Storage | Storage |
| Azure Function | Serverless |
### DigitalOcean
| Type | Category |
| ---------------- | ---------- |
| Droplet | Compute |
| Managed Database | Database |
| DOKS Cluster | Kubernetes |
| Space | Storage |
| Domain | DNS |
| Load Balancer | Network |
## Viewing Resources
Navigate to **Resources** in the sidebar to see your full cloud inventory. The page displays:
* **Summary cards** at the top showing total resource count, EC2 instances, RDS databases, and running resources
* **Filter bar** with six filters that can be combined
* **Paginated table** listing all matching resources
### Available Filters
| Filter | Description |
| ------------ | -------------------------------------------------------------------------- |
| **Search** | Free-text search by resource name or ID |
| **Provider** | Filter by cloud provider (AWS, GCP, Azure, DigitalOcean) |
| **Account** | Filter by a specific cloud account |
| **Type** | Filter by resource type — automatically scoped when a provider is selected |
| **Status** | Filter by resource status (Running, Available, Stopped, Pending, Error) |
| **Region** | Filter by cloud region — populated from your discovered resources |
Select a **Provider** first to narrow the **Type** dropdown to only the resource types available for that provider.
Use the **Refresh** button in the filter bar to reload the current view without changing your filters. To remove all filters at once, click **Clear all**.
## Resource Statuses
| Status | Meaning |
| -------------- | -------------------------------------------------------------------------------- |
| **Running** | The resource is active and operational |
| **Available** | The resource is provisioned and ready (common for databases and storage) |
| **Stopped** | The resource exists but is not running (e.g., a stopped EC2 instance) |
| **Pending** | The resource is transitioning between states (starting, stopping, etc.) |
| **Error** | The resource is in a failed or unhealthy state |
| **Terminated** | The resource has been destroyed — it will be removed on the next discovery cycle |
## Compute Management
Upmetr allows you to **start**, **stop**, and **reboot** EC2 instances directly from the dashboard.
Compute management actions are restricted to **Superadmin** users and require the **compute\_management** module on your plan. These actions affect live infrastructure — use with caution.
### Available Actions
| Action | Required Instance State | Result |
| ---------- | ----------------------- | --------------------------------- |
| **Start** | Stopped | Instance transitions to Running |
| **Stop** | Running | Instance transitions to Stopped |
| **Reboot** | Running | Instance restarts (stays Running) |
### How It Works
1. Navigate to a resource detail page for an EC2 instance
2. Click the desired action button (**Start**, **Stop**, or **Reboot**)
3. A **confirmation modal** appears — you must explicitly confirm the action
4. Upmetr executes the action via the AWS API and updates the resource status
5. The action is recorded in the **Audit Log** for security tracking
Compute management currently supports **AWS EC2 instances** only. Your IAM policy must include `ec2:StartInstances`, `ec2:StopInstances`, and `ec2:RebootInstances` permissions. See the [Cloud Accounts guide](/guides/cloud-accounts) for IAM policy details.
## Manual Discovery
If you need to refresh your resource inventory immediately (for example, after launching new instances), you can trigger a manual discovery:
1. Go to **Settings > Cloud Accounts**
2. Click the **Discover** button on the account you want to re-scan
3. Upmetr scans the account and updates the resource list
The next automatic discovery will continue on its regular 15-minute schedule.
## Troubleshooting
* Ensure you clicked **Discover** on the cloud account after adding it, or wait up to 15 minutes for automatic discovery
* Verify that the credentials have the correct read permissions for the resource types you expect (e.g., `ec2:Describe*`, `rds:Describe*` for AWS)
* Check **Settings > Cloud Accounts** and run **Test Connection** to confirm the account is connected
* Resources are refreshed every 15 minutes automatically
* Click the **Refresh** button on the Resources page to reload the current data from the database
* To force a full re-scan from the cloud provider, go to **Settings > Cloud Accounts** and click **Discover**
* Verify the IAM policy or service account role includes permissions for the missing resource type
* Some resource types are region-specific — make sure the account's default region or multi-region scanning covers the expected regions
* Check that the resource exists and is not in a terminated state in your cloud provider's console
* Compute management requires the **compute\_management** module — check your plan includes it under **Settings > Billing**
* Only **Superadmin** users can perform start/stop/reboot actions
* The IAM policy must include `ec2:StartInstances`, `ec2:StopInstances`, and `ec2:RebootInstances` in addition to read-only permissions