Data Connections and Output
Connection and output questions test reliability and correctness of I/O. Input Data supports files, databases (via ODBC/OLEDB), and Alteryx data connections. Output Data writes files, databases, and can append, overwrite, or create. The Dynamic Output tool writes to paths supplied by a field (one file per group). Data Stream Out/In passes data between workflows. Render and Layout produce formatted reports. Download retrieves files. Publish to Gallery shares workflows and apps.
Key concepts: file formats and their traps - CSV loses types (leading zeros, dates), YXDB preserves types and is fast, Excel has row limits and sheet naming rules, Parquet preserves schema. Database writes need correct field mapping and permissions; PreSQL/PostSQL run statements around the write. Output options: Overwrite (replace file), Append to existing, Create new sheet. When writing to a database, the tool must match column types or fail.
Traps: Output Data to a CSV then re-reading changes field types; Dynamic Output requires a unique path per record or files overwrite each other. Writing to a file already open (Excel) fails. Connection credentials stored in a workflow may not travel - use saved connections. Block Until Done is often needed before reading a file you just wrote.
Worked example: split a dataset into one CSV per region. Use Formula to build "C:\out\" + [Region] + ".csv", then Dynamic Output with that path field. Rule of thumb: Use YXDB for intermediate storage to preserve types; use CSV only for interchange.