1. Docs
  2. Resources
  3. Data Source

Data Source

Data Source is where a project keeps its test data and its connections to outside systems: named values (keys), Excel or CSV datasets, and database and Kafka connections. Test steps read them by name, so a value or a password changes in one place.

6 min read Updated CustomersAdmins

In short

On this page you can

  • Keep test data as named keys that steps read with {{ds:key}}.
  • Store passwords and tokens encrypted so nobody sees them again.
  • Upload Excel or CSV Datasets for data-driven runs.
  • Define database and Kafka connections once for every DB and Kafka step.
  • Give a key a different value per environment with _DEV, _TEST and _PRP.

How to use it

  1. Open Data Source in the main menu and pick a tab.
  2. In Variables, click New key, name it and set its value.
  3. In Connections, add a DB or Kafka connection and click Test Connection.
  4. In Automate, insert the key into a step with the data source picker icon.
  5. Run the test; if a value is wrong, check the step log and the key’s name.

At a glance

Where
Main menu › Data Source. Tabs: Variables, Datasets, Connections, each with a count.
Who
Every project member creates and edits. Deleting a key, or a folder that is not empty, needs an Admin. Only the creator of an encrypted value can show it.
You need
A selected project. For DB and Kafka: the server address and a test user that can log in.
Time
A key takes seconds. A tested database connection takes a few minutes.
Produces
Keys ({{ds:key}}), datasets ({{excel:name}}), DB connections and Kafka connections.
Used by
Automate steps (any parameter, API, DB, Kafka), pipelines that loop over a dataset, and the AI (key names only, never values).
Project types
The same in WEB, MOBILE and DESKTOP projects. DB and Kafka connections are opened from the machine that runs the test, so the server must be reachable from there.

Start from your question#

Where Data Source fits#

Data Source sits beside the order of work. You fill it before or while you automate, and every later step reads from it.

  1. 1DiscoverConnected actions tell the AI which pages need which keys
  2. 2AnalyzeNo direct use
  3. 3AutomateSteps read keys, cells and connections
  4. 4SchedulePipelines loop over datasets and pick the environment
  5. 5ReportEncrypted values are masked in logs
  6. 6MaintenanceChange a value once, every test follows

Why use it#

  • One value, many tests. A step stores {{ds:loginEmail}}, not the email itself. Change the key and every test that uses it picks up the new value on its next run.
  • Secrets stay secret. An encrypted key is stored encrypted, never shown again on any screen, and masked as •••••• in run logs and reports. See Secrets and credentials.
  • One test, several environments. Add apiBase_TEST next to apiBase and a run on TEST uses it automatically. See Environment-specific keys.
  • Live data instead of hard-coded data. A key can take its value from an SQL query at run time, or from a cell of a dataset.
  • Connections defined once. Every DB and Kafka step picks a connection by name. The password lives in the connection, not in the step.

The building blocks#

Key
A named value, for example loginEmail = qa.user@example.com. Unique in the project, no spaces. Steps read it as {{ds:loginEmail}}. The Variables tab lists all keys.
Value source
Where a key’s value comes from: Typed value, Dataset cell or DB query. See Where a value comes from.
Encrypted key
A key whose value is stored encrypted. Only a test run reads it. Use it for every password, token and API key.
Environment-specific key
A key whose name ends with _DEV, _TEST or _PRP. A run in that environment uses it instead of the key without the suffix.
Dataset
A table uploaded from XLSX, XLS or CSV, or typed in. Steps read a column as {{excel:name}}_column; a pipeline can run once per row.
DB connection
A database login (JDBC URL, driver class, user, password) that DB steps and DB-bound keys use.
Kafka connection
A broker login (bootstrap servers, security protocol, SASL user and password) that Kafka Produce and Consume steps use.
Connected actions
The Discover actions a key belongs to, for example the email field of the login page. They tell the AI which pages need which data. Values are never sent.
Folder
Groups keys, datasets or connections. Folders change nothing at run time.

The screen#

The page opens on the tab in the address (?tab=tables for Datasets, ?tab=connections), otherwise on Variables. An open key, dataset or connection is kept in the address too, so you can share a link to it or refresh without losing it.

TabWhat it holdsGo there when
VariablesEvery key, grouped by folder, with its kind, default value, environment and connected actions.You add or change test data or a secret. Variables
DatasetsExcel and CSV tables with their size, columns and the keys that read them.You want data-driven runs or a key that reads a cell. Datasets
ConnectionsDatabase and Kafka connections, with the keys that use each database.A step must query a database or send a Kafka message. Connections

The test mailbox for Mail steps is not here. It is set up under Integrations (Email).

Who can do what#

TaskUserAdmin
Create and edit keys, datasets, connections and foldersYesYes
Delete a key, or a key folder that still has contentNoYes
Delete an empty folder, a dataset or a connectionYesYes
Show an encrypted valueOnly its creatorOnly its creator
Replace an encrypted valueYesYes

Roles are explained in Roles & permissions. Everything belongs to one project; there is no sharing of keys or connections between projects.

Before you start#

  • Select a project. Without one the page is read-only.
  • Use test data. Values and dataset rows are visible to every project member, except encrypted values.
  • For a database: the JDBC URL, the driver class, and a test user with only the rights your tests need.
  • For Kafka: the bootstrap servers and, for SASL, the mechanism, user and password.
  • Network: the database or broker must accept connections from SmartestQA (for Test Connection) and from the machine that runs the tests: the SmartestQA server for Run on Server, or your agent. See Where connections are opened from and Do you need an agent?

Best practices#

Do

  • Encrypt every password, token and API key.
  • Name keys the way testers talk: loginEmail, apiBase.
  • Put environment differences in _DEV, _TEST, _PRP keys, not in separate tests.
  • Click Test Connection before the first save of a connection.
  • Give DB connections a read-only user unless a test must change data.
  • Fill Description so others know what a key is for.

Don’t

  • Type a password into a test step. Use an encrypted key.
  • Store real customer data in datasets or keys.
  • Point a DB connection at production with a user that can write.
  • Delete a key before you have checked which tests use it.
  • Use one personal mailbox or login for automated tests.

In this section#

Didn’t find your answer?Write to support with your project name and what you see on screen.

info@smartestqa.ai
© 2026 Selenity Technology · SmartestQAsmartestqa.aiThis page as MarkdownChecked against smarttest-fe predev 2a8c79cc, smarttest-be 633c2a03, smarttest-testsuite (runner), smartest-mcp-server