Sink Node
TLDR: Sink nodes are endpoints where agents leave your simulation. Choose whether each sink counts as a successful completion or as waste/defects — this affects your analytics and yield calculations.
What Happens Here?
The Sink properties panel shows: "How should agents counted here be categorized?"
You see two toggle cards:
| Option | Behavior | Use When |
|---|---|---|
| Completed Successfully | Agents counted as successful completions in analytics | Customers who finished, products that passed QC, patients discharged |
| Defective / Waste | Agents counted as scrap in analytics | Failed QC items, abandoned customers, timed-out orders |
The waste flag affects throughput and yield calculations. Successful completions count toward throughput; waste exits count toward defect rate. This lets you calculate metrics like first-pass yield without extra setup.
Multiple Exits
Most simulations benefit from more than one Sink node. Common patterns:
- Main sink (Completed Successfully) — the happy path
- Waste sink (Defective/Waste) — failed inspections or scrapped items
- Balking sink (Defective/Waste) — customers who left due to long waits
- Timeout sink (Defective/Waste) — items that exceeded time limits
Label each Sink descriptively (e.g., "Completed Orders", "Failed QC", "Abandoned") so your simulation results are easy to read.
Not every "bad" outcome needs a Sink. For rework scenarios, connect a path from a quality check back to an earlier Processor instead. Only route to a waste Sink when items are truly scrapped.