Skip to main content
In ubx, every cloud resource is declared with a unit block. Each unit maps 1:1 to a Pulumi provider resource.

​
Resource Addresses

Resources are addressed using dot notation: 
unit.resource_type.name
Examples:
  • unit.aws_s3_bucket_v2.assets
  • unit.aws_rds_instance.db
  • unit.google_storage_bucket.media

​
Resource Outputs

Every cloud resource has output attributes — values known only after the resource is created. Access them with the ~ prefix: 
~unit.aws_rds_instance.db.endpoint   # Pending<string>
~unit.aws_s3_bucket_v2.assets.arn    # Pending<string>
~unit.aws_eks_cluster.cluster.name   # Pending<string>

​
Providers and Resource Types

ubx uses the same resource type names as Terraform/Pulumi, in snake_case:
ProviderExample type
AWSaws_rds_instance, aws_s3_bucket_v2, aws_vpc
GCPgoogle_storage_bucket, google_sql_database_instance
Azureazurerm_resource_group, azurerm_storage_account
KubernetesUsed via deploy and sync blocks
Run ubx schema list to see all registered types. Run ubx docs <type> --list to see fields.

​
Viewing Resource Schema

ubx docs aws_rds_instance --list
# Required:
#   engine               string
#   instance_class       string
#   ...
# Optional:
#   allocated_storage    number
#   multi_az             bool
#   ...

​
Resource Lifecycle

Resources go through these states:
  1. Planned — ubx plan shows what will happen
  2. Created — ubx apply provisions the resource
  3. Updated — ubx apply after attribute changes
  4. Refreshed — ubx refresh syncs state with actual cloud
  5. Destroyed — ubx destroy removes the resource