Add Resources from Catalog#

Resources are the individual cloud service instances you provision through Reply CMP—for example, an Azure Virtual Network, an AWS S3 Bucket, or a GCP Cloud SQL database. Resources live inside a Deployment: the Deployment provides the connections, the Terraform state, and the apply lifecycle.

Note

You must have an existing Deployment before you can add Resources. To create one, see Manage Deployments.

Navigate: left navigation → Provisioning → open a Deployment → Add Resource (or click “Deploy from Catalog” on the Deployment detail page).

Step 1 — Choose a catalog item#

Browse available service templates. Each tile shows:

  • Provider badge (Azure / AWS / GCP)

  • Template name (for example, “Azure Kubernetes Service”, “AWS S3 Bucket”)

  • Short description

  • Required fields preview

Use the search bar to filter by name. Use the provider filter to narrow by cloud. Select a template to proceed.

Note

The template catalog is managed by your platform team. If a resource type you need is not available, contact your administrator to request it.

Step 2 — Resource details#

Field

Description

Resource name

Max 40 characters; must be unique within the Deployment

Connection alias

Which of the Deployment’s attached connections this Resource targets (for example, default, prod, dev). The dropdown shows only aliases already attached to the Deployment.

Region

Provider region (choices depend on the selected catalog item)

Note

Connection aliases are set at the Deployment level when the Deployment is created or edited. If the alias you need is not available in the dropdown, add the corresponding Connection to the Deployment first.

Click Next to continue.

Step 3 — Configuration#

A dynamically generated form based on the selected catalog item’s JSON Schema. The form validates inline as you type.

Feature

Description

Required fields

Marked with an orange asterisk; form cannot advance until complete

Completion progress bar

Shows % of required fields filled; reaches 100% before enabling Next

Inline validation

Errors appear per-field in real time — no need to submit to discover them

JSON toggle

Click “View as JSON” to switch to a raw JSON editor

Default values

Some fields are pre-filled; review before modifying

Tip

Use the JSON view toggle to paste a configuration from a previous Resource or an external source. The form and JSON view stay in sync.

Click Save to save the Resource. The Resource is now in Draft status — the cloud has not changed yet.

Resource statuses#

Status

Meaning

Draft

Resource metadata has been saved or modified but not yet applied; the live cloud resource is unchanged

Valid

Resource was successfully applied; latestAppliedMetadata matches the current metadata

Note

These are Resource statuses. The Deployment has its own separate status (Valid, Error, Tainted). A Deployment can be Valid while some of its Resources are in Draft, waiting for the next apply.

Step 4 — Applying your changes#

Resources in Draft are not provisioned until you run an Apply on the Deployment. Apply and Dry run operate on all Resources in the Deployment at once.

Dry run (plan)#

From the Deployment detail page, click “Run Dry Run” to preview what will change without making any modifications:

  • A terminal-style output pane shows the Terraform plan.

  • CREATE / MODIFY / DESTROY counts are displayed.

  • An AI Plan Summary appears below the terminal with a plain-language description of what will happen.

Review the summary and the plan output. If everything looks correct, click Apply.

Apply#

The platform runs terraform apply for the whole Deployment. Progress is shown in the terminal pane. When complete:

  • Success: An AI Plan Summary (post-apply) describes what was created; a link to the new resource in Discovery is shown. The Resource status changes to Valid and latestAppliedMetadata is updated. The Deployment returns to Valid.

  • Failure: An AI Error Explanation describes in plain language what went wrong and suggests corrective actions; the raw Terraform output remains available for investigation. The Deployment enters Error status; Resource metadata is unchanged.

Note

The AI Plan Summary and Error Explanation are generated by Azure OpenAI and may not be 100% accurate in edge cases. Always cross-check with the raw Terraform output if you are unsure.

Editing a resource#

  1. Open the Deployment → find the Resource → click Edit.

  2. The configuration form (Step 3) re-opens with the current values.

  3. Make changes — inline validation applies.

  4. Click Save → the Resource moves back to Draft status. The live cloud resource is not changed yet.

  5. Run Dry run and then Apply from the Deployment to push the changes.

Tip

To speed up creating similar Resources, duplicate an existing one and only adjust the parameters that differ (for example, name, tags, or size).

Resource activity log and rollback#

The History tab on a Resource detail page shows:

  • A timeline of every metadata save (with timestamp and user).

  • A diff of what changed at each version.

  • A Restore button on each entry — click to load that version’s configuration back into the form.

After restoring a version, the Resource moves to Draft. Run a Dry run and Apply to confirm and push the restored configuration.

Deleting a resource#

Deleting a Resource in the UI schedules its destruction—it does not immediately remove the cloud object. The actual deletion occurs on the next Apply of the Deployment, when Terraform removes the corresponding infrastructure.

  1. Open the Resource → click Delete → confirm.

  2. The Resource is marked for deletion.

  3. Run Apply on the Deployment → Terraform destroys the cloud object.

  4. The Resource is removed from the Deployment.

Warning

If the Deployment has no Resources left after deletion, the next Apply will leave the Deployment in Tainted status (eligible for Deployment deletion). See Manage Deployments.

Connection aliases#

Resources target a specific Connection using its alias. Aliases are defined at the Deployment level:

Alias

Typical meaning

default

The first Connection added for a given provider type

prod

Production account

dev

Development / sandbox account

custom

Set by the administrator when creating or editing the Deployment

Select the correct alias carefully — the Resource deploys to the cloud account associated with that alias.