Processor Node
TLDR: Processor nodes are service stations where work happens. Configure how many agents can be served at once, how long each takes, how the queue is managed, what staff are required, and when the station is open.
The Work
The top section of the Processor properties panel has two parts:
Capacity
You see: "Process up to [number] agents at a time"
This is how many agents can be served simultaneously. Set it to 1 for a single server (one barista), or higher for parallel servers (3 bank tellers sharing one queue).
A single Processor with capacity 3 is different from three separate Processors. One queue feeding 3 servers is more efficient — agents go to whichever server finishes first. Three separate Processors each have their own queue, so agents might wait in a long line while another sits empty.
Service Time
You see: "Each takes about [number] ± [variation] [seconds / minutes / hours]"
The first number is the average time to process one agent. The ± value adds natural variation.
Below the timing card, a link reads "Have real data? Fit my timing" — click it to open the Distribution Detective and automatically fit your historical service time data.
For full details on timing modes, see Time & Distributions.
Resource Requirements
Only visible when 2 or more Resource pools are connected.
When a Processor needs staff from shared resource pools, the Staffing Manifest appears. It lets you build complex staffing rules:
- REQUIRE ALL (AND) — the Processor needs one resource from every pool in the group
- REQUIRE ANY (OR) — the Processor needs one resource from any pool in the group
Groups combine with AND logic between them. For example:
(Nurse OR Phlebotomist) AND (Room) — needs any one nurse-type worker, plus a room.
Click resource pool chips to toggle them in or out of a group. Use "Add Staffing Group" to create additional groups.
Line Priority
Controls the order agents are pulled from the queue. You see a set of preset cards:
| Preset | Behavior | Use When |
|---|---|---|
| Fair Line | First come, first served | Default fairness (most common) |
| Most Recent First | Last in, first out | Manufacturing buffers, stacks |
| Longest Waiting First | Sort by wait time (ascending) | Service level guarantees |
| Shortest Wait First | Sort by wait time (descending) | Quick-service optimization |
| Custom Priority Rules | Rank by any agent attribute | ER triage, VIP customers |
If your simulation has agent attributes (like triage color or ESI level), additional context-aware presets may appear.
Custom Priority
When you select Custom Priority Rules, you choose an attribute to rank by:
- Numeric attributes: Choose "Lowest value first" or "Highest value first"
- Categorical attributes: Drag values into your preferred order (e.g., Red > Orange > Yellow > Green)
Agents Left Without Processing
Only visible when 2 or more outgoing connections exist.
This section models balking — agents who leave the queue before being served. You configure:
- Where do they go? — a dropdown to pick which outgoing connection leads to the balking destination (typically a waste/defect Exit)
- Max wait time: "Leave if waiting longer than [number] [unit]" — agents abandon the queue after this time. Set to 0 for no limit.
Shifts & Breaks
A Gantt-style timeline where you define when the Processor is operational. Add shift blocks to set working hours. Outside of shifts, the Processor stops accepting new agents (agents already being processed will finish).
Use this to model lunch closures, shift changes, or part-time schedules.
Path Control
Only visible when 2 or more outgoing connections exist.
Same routing options as the Entry node: Random, Custom Split, Shortest Queue, Priority, or Conditional. See Entry Node for details.
Combine an Entry node using Schedule mode (high arrivals at noon, low in the afternoon) with a Processor using Shifts (closed 2–3pm for cleaning). This captures realistic restaurant dynamics.