NoopOperator

Pass-through operator that forwards input data unchanged. Short name: noop · Category: Functional

Overview

The NoopOperator copies its input table to output without modification, optionally sleeping for a configurable duration. Use it to test flow structure before implementing real operators, to introduce timing delays for performance testing, or to isolate a problem between two operators by inserting a no-op between them.

Key Features

• Passes all rows and columns through unchanged
• Adds no new columns to the PyArrow table
• Configurable sleep delay for simulating slow operators
• Zero performance overhead when sleep_sec is 0
• Safe for use in any position in a flow

Operator Configuration

{
  "name": "noop_step",
  "type": "noop",
  "depends_on": ["previous_op"],
  "config": {
    "sleep_sec": 0
  }
}

Parameters

Output Columns

This operator adds no columns. The output table is identical to the input table.

Examples

Example 1: No-op between two operators (debugging)

{
  "name": "debug_checkpoint",
  "type": "noop",
  "depends_on": ["extract"],
  "config": { "sleep_sec": 0 }
}

Example 2: Simulate a slow operator (performance testing)

{
  "name": "slow_step",
  "type": "noop",
  "depends_on": ["ingest"],
  "config": { "sleep_sec": 5 }
}

Troubleshooting

Flow completes instantly but you expected a delay — ensure sleep_sec is set to a positive integer, not a float.

Using noop in production — remove noop operators before production deployment. They add no value and inflate flow execution time when sleep_sec > 0.

Sample Flow

See sample_flows/advanced/quality_branching_merge_pipeline.json for a complete pipeline example.