Stream Configuration Overview

Use this page when you need to understand how a stream configuration is organized before looking up individual field paths.

A stream configuration defines:

what to read in source
where to write in target
how to control execution with mode, reportingInterval, and optional limits

Saved configurations add metadata such as id and created, but the working shape stays the same.

Core structure

Every stream configuration follows the same top-level pattern:

{
  "name": "My stream",
  "mode": "load",
  "reportingInterval": 5,
  "source": {
    "connections": [],
    "options": {}
  },
  "target": {
    "id": "conn_target",
    "spec": {}
  },
  "limits": {}
}

Mental model:

source.connections[] is always an array, even for a single source
each source connection chooses one source family: database fields, s3, or files
target.spec always contains exactly one target family: db, files, or s3
limits is optional and is mainly used for bounded runs and validation

Annotated example

{
  "name": "PostgreSQL to MySQL",
  "mode": "load",
  "reportingInterval": 3,
  "source": {
    "connections": [
      {
        "connectionId": "conn_postgres_prod",
        "database": "postgres",
        "tables": [
          { "name": "public.actor" }
        ]
      }
    ],
    "options": {
      "dataBundleSize": 1000
    }
  },
  "target": {
    "id": "conn_mysql_target",
    "spec": {
      "db": {
        "database": "testdb",
        "schemaPolicy": "create_missing_only",
        "writeMode": "upsert"
      }
    }
  }
}

Read it as:

name, mode, and reportingInterval define the run at the top level
source.connections[0] selects one stored source connection and scopes what to read from it
source.options adds execution settings that belong to the read side
target.id selects the destination connection
target.spec.db tells DBConvert Streams to write into a database target

Source vs target

source and target are intentionally asymmetric:

source.connections[] can contain one or more connections
each source connection carries its own scope, such as selected tables, a manifest path, or local file paths
target always points to one destination connection at a time
target.spec decides how that destination should be written to

Which page to open next

Use Source Configuration when you need to define tables, files, queries, or S3 scope.
Use Target Configuration when you need to define database writes, file formats, staging, or upload behavior.
Use Stream Configuration Reference when you need exact field paths.
Use Stream Configuration Examples when you want full copyable payloads.

Verification

Before saving or using a payload:

confirm it uses source.connections[] rather than an old flat source.id shape
confirm each source connection uses only one source family
confirm target.spec contains exactly one family object
confirm saved configs add metadata such as id and created, but your working payload does not depend on those fields