Home› Getting Started› Quick Start Guide› Create Connectors

Create Connectors

After initializing the repository, the next step is to define connectors to source systems. Connectors provide the technical and structural foundation for importing metadata and generating the data warehouse model.

A connector defines how AnalyticsCreator accesses a source system and how metadata (tables, columns, keys, relationships) is retrieved. This metadata is then stored in the repository and used by the data warehouse wizard to generate a draft model.

Purpose

Establish access to source systems and import structural metadata required for automated data warehouse generation.

Design Principle

AnalyticsCreator separates metadata acquisition from data extraction.

  • Metadata (structure) is imported first and stored in the repository
  • Data extraction happens later during execution (via pipelines)

This means a data warehouse model can be designed and generated without requiring an active connection to the source system at runtime.

Inputs / Outputs

Inputs

  • Connector type (e.g. SQL Server, SAP, metadata connector)
  • Connection configuration (server, database, authentication)
  • Selected schemas, tables, or metadata source

Outputs

  • Connector definition stored in repository
  • Imported metadata:
    • Tables and views
    • Columns and data types
    • Primary keys
    • Foreign keys or references (if available)

Internal Mechanics

1. Connector definition

The connector stores the configuration required to access a source system. This includes connection details and selection of relevant schemas or objects.

2. Metadata extraction

AnalyticsCreator reads structural metadata from the source system or from a metadata connector. This includes:

  • Table structures
  • Column definitions
  • Key definitions
  • Relationships between tables

In some cases (e.g. SAP or metadata connectors), metadata can be imported without direct access to the operational system.

3. Repository persistence

All imported metadata is stored in the repository. At this stage:

  • No SQL objects are generated
  • No data is extracted
  • No pipelines are executed

The system builds a structural model that will later drive code generation.

4. Relationship availability

If source systems expose foreign keys or references, these are imported and can be reused during modeling. If not, relationships must be defined manually in later steps.

Types / Variants

Connector types

  • Direct database connectors (e.g. SQL Server)
  • ERP connectors (e.g. SAP metadata extraction)
  • Metadata connectors (predefined structures without live connection)

Import modes

  • Full metadata import
  • Selective table import
  • Manual definition (if metadata is incomplete)

Example

A connector is created for a SQL Server database containing the following tables:

  • Customer
  • Orders
  • OrderLines

The system imports:

  • Column definitions (e.g. CustomerID, OrderID)
  • Primary keys
  • Foreign key relationships (e.g. Orders → Customer)

These definitions are stored in the repository and become available for automated model generation in the next step.

When to Use / When NOT to Use

Use when

  • Starting a new data warehouse model
  • Importing metadata from source systems
  • Preparing for automated model generation

Do NOT rely on connectors alone when

  • Source metadata is incomplete or inconsistent
  • Business relationships differ from technical relationships
  • Required structures are not exposed in the source system

Performance & Design Considerations

  • Connector scope directly affects model complexity
  • Importing unnecessary tables increases modeling overhead
  • Metadata quality determines quality of generated model

Design trade-off:

  • Broad import (high coverage, more noise)
  • Selective import (cleaner model, more manual work later)

Integration with other AnalyticsCreator features

  • Repository: stores connector and metadata definitions
  • Wizard: uses imported metadata to generate draft model
  • STG generation: based on imported source structures
  • Transformations: reuse source metadata and relationships

Common Pitfalls

  • Importing entire source systems without filtering
  • Assuming source relationships are suitable for analytical models
  • Using technical keys as business keys without validation
  • Skipping metadata validation before running the wizard

Key Takeaway

Connectors import and persist source metadata in the repository, forming the structural basis for automated data warehouse generation.