Docs/Core concepts
Docs/Core concepts

Source systems

A source represents a data system that your Vaurd flows validate against: a database, data warehouse or ERP. A flow reads the data it needs from a source. You can create as many sources as your organization needs.

A source is a metadata definition, not a live connection. It describes what the system is; the credentials used to connect to it are configured separately in the agent, inside your own environment.

Scope

Sources are defined at the organization level and are shared across all workspaces in your organization. Any flow, in any workspace, can use any source your organization has defined. Sources belong to your organization and are not visible to other organizations; each organization defines and manages its own sources.

Supported source types

Databases

SQL Server, MySQL, PostgreSQL.

Data warehouses

Databricks, Snowflake, Google BigQuery, Athena.

ERP systems

Via Dataverse (for Microsoft ERP products).

What a source contains

You create a source from the Vaurd web application. When you create a source, you define:

Name
The name of the source.
Technical Name
The identifier used to reference the source, which will be defined in your agent's configuration.
Description
A description of the system.
Source type
One of the supported types listed above.

Creating a source

Sources are created in the Vaurd web application, under Configure → Sources.

  1. Open Configure → Sources and select New Source.
  2. Enter a Name (the display name) and a Technical Name (used to reference the source in your agent's configuration).
  3. Select a Source type from the supported types.
  4. (Optional) Add a Description.
  5. Select Create.
Screenshot: the 'Create new source' dialog, with a PostgreSQL source type selected, Source name (Orders DB), Technical name (orders_db), and an optional Description (Primary database for customer orders).

Credentials are not entered here. Creating a source only registers its definition in the platform; you supply the connection details separately in the agent configuration (see below).

Configuring source credentials in the agent

Because the agent runs inside your environment and connects to your systems locally, the agent (not the platform) holds your source credentials. You provide them in the agent's config.yaml file, under the sources section.

Each entry maps a source to its connection details:

name
The source name.
type
The source type, (for example database_mysql).
params
The connection parameters for that source type (host, port, user, password, database, and so on).

For example:

config.yamlYAML
sources:
  - name: mysql-source    # should reference technical name of your source defined in platform
    type: database_mysql
    params:
      host: localhost
      port: 3306
      user: root
      password: my-secret-pw
      database: testdb

At runtime, the agent uses these credentials to connect to the source and fetch the data your flows need. Because credentials live only in the agent config within your environment, Vaurd never stores or transmits them.

Using sources in flows

To read from a source, a flow uses a Fetch node, which selects the source and queries it. See the Fetch node for details.

Diagram: inside your environment (VPC / on-prem), the Vaurd agent's Fetch node reads credentials from config.yaml and uses them to connect directly to the source system database. Credentials stay in the agent, inside your environment.