Data sources
Two tabs configure where a Flow reads and writes tabular data: SQL Connections for a database, and CSV Configs for files the Robot writes itself.
SQL Connections
An SQL Connection points at a database a Flow can query.
| Field | Meaning |
|---|---|
| Connection name | How the connection is identified when selecting it in a Node |
| Engine | PostgreSQL, MySQL, MariaDB or MSSQL |
| Host | Hostname or IP of the database server |
| Port | The server's port |
| Username / Password | The credentials to connect with |
| Database | Which database on that server |
Use Add Config to create one.
Test connection
Test connection connects with the credentials as entered and reports the result. On success
it also lists the table schemas it found — every table with its columns, each column's type,
and whether it is NOT NULL.
Use it before writing the query rather than after: the schema listing tells you the exact column names and types, which is where most query problems actually come from. A successful connection to a database with no tables says so explicitly, which distinguishes "wrong database" from "wrong credentials".
Queries themselves are written in the SQL Query Node.
Credentials are stored on the Control Unit as part of its configuration, and they travel with an exported configuration. Use an account scoped to what the Flow actually needs — not an administrative one — and review exports before sharing them.
CSV Configs
A CSV Config defines a file the Robot writes: which Variables become columns, in what order.
| Field | Meaning |
|---|---|
| Config Name | How the config is identified in a Node and in the CSV Download Widget |
| Columns | The Variables logged, one column per Variable |
Use Add Variables to pick the Variables — the picker has a search field, select-all and clear, which matters on a Robot with a large Variable set.
Rows are appended from a Flow with the CSV Node.
Preview and snapshots
The config shows a preview of the current file — the most recent rows, with a count of how many are shown out of the total. This is the quickest way to confirm a Flow is logging what you expect, and in the right column order.
RobFlow also keeps snapshots of previous files. You can select a snapshot to preview it, and download either the current file or Download all, which produces the latest file plus every backup as a ZIP.
Deleted Variables
If a Variable used as a column is deleted, the config keeps the column but marks it as
(deleted). The config keeps working — the column is simply no longer populated.
This is deliberate, so deleting a Variable does not destroy the historical file. But a config quietly logging an empty column is easy to overlook: check the Columns list after any Variable cleanup, and remove the dead entries.
Getting the data out
- On a Dashboard — the CSV Download Widget serves a chosen CSV config to an operator, so they can pull a shift's data without opening RobFlow. See Misc Widgets.
- To a database — the SQL Query Node writes directly, which suits a line where an MES already owns the data.
- To another system — the HTTP Request Node posts wherever you need.