Connections

What is a connection?

A connection describes a single infrastructure endpoint where data in your product can be found. It has a name, a type, and a host — and optionally a set of key-value variables for any additional configuration consumers might need.

Connections are declared once in the spec and referenced by ports. This separation means you can update an endpoint (say, a database host) in one place without touching every port that uses it.

Defining a connection

Add connections interactively:

daco connections add

Or pass all values as flags:

daco connections add \
  --name analytics_db \
  --type postgresql \
  --host analytics.db.example.com \
  --description "Production analytics warehouse"

In your opendpi.yaml, this produces:

connections:
  analytics_db:
    type: postgresql
    host: analytics.db.example.com
    description: Production analytics warehouse

Connection types

The type field is a free-form string. There is no enumeration — use whatever label clearly communicates the technology to your team and downstream tools.

Common values:

Type	Typical host format
postgresql	hostname or hostname:port
mysql	hostname or hostname:port
kafka	broker:port
s3	s3.amazonaws.com or region endpoint
http	base URL (e.g., api.example.com)
bigquery	GCP project ID
snowflake	account identifier
databricks	workspace hostname

Variables

Use variables for any extra configuration that consumers need to connect — database names, schema names, bucket names, prefixes, etc.

connections:
  warehouse:
    type: postgresql
    host: dwh.example.com
    variables:
      database: production
      schema: public
      port: "5432"

Variables are always string values. They appear in daco connections describe output and are available to downstream tooling that reads the spec.

Referencing a connection from a port

When you add a port and bind it to a connection, the spec stores a JSON Reference:

ports:
  daily_metrics:
    connections:
      - connection: "#/connections/analytics_db"
        location: customer_daily_metrics

The connection value is always "#/connections/<name>". The CLI generates this automatically — you should not need to write it by hand.

Listing and inspecting connections

daco connections list

Shows all connections, their types, hosts, and how many ports reference each one.

daco connections describe analytics_db

Shows the full connection detail, including variables and the list of ports that use it.

Removing a connection

daco connections remove analytics_db

Connections that are referenced by one or more ports cannot be removed. Remove or rebind those ports first, then delete the connection.

Use --force to skip the confirmation prompt:

daco connections remove unused_conn --force

Multiple connections per port

A port can reference more than one connection — for example, if the same data is mirrored to two stores, or if the port spans a primary source and a replica:

ports:
  events:
    connections:
      - connection: "#/connections/kafka_prod"
        location: customer-events
      - connection: "#/connections/s3_archive"
        location: events/customer/