Purpose
This block starts other flows as separate executions during the write stage. Use it for a loading, recalculation and notification chain, or processes that read and modify external systems. For pure transformation returning a table, use the separate “Call flow” subflow block.
Dependency and result
The input is optional; when connected, the call follows the preceding step. Input data can be passed to child flows. Success and failure outputs describe the call outcome. Returning a child’s transformed table is not this block’s purpose.
Flows and values
- Choose up to 20 flows and arrange their order. Their Production versions are used.
- For each, enable parameter passing when needed: constant, current launch parameter or first-row field.
- “Pass block data” sends all fields or a selected list.
- “Wait for completion” is enabled by default. Without waiting, acceptance of the run is confirmed rather than its business result.
Configuring a chain
- Publish each child flow and verify the initiating user’s execution rights.
- Add “Receive data from flow” in a child that needs an input table.
- Choose flows and parameter sources; ensure first-row values are appropriate.
- Set data passing and completion waiting.
- Simulate the parent, then inspect actual child runs using a sample dataset.
Background execution
Child flows run unattended like scheduled jobs; a wizard does not request values. Required parameters must be supplied by the call or have defaults. Simulation validates configuration without starting children. Call cycles are prohibited and maximum chain depth is 5. The user needs access to each flow, not only the parent.
Post-load example
After a successful actual-data write, call “Recalculate measures” and pass Period. Wait for completion before continuing to the finance-team notification. If recalculation performs its own writes, inspect its separate log and do not confuse run acceptance with completed recalculation.