Flow Diversion / Rerouting¶

Status: draft
Scope: cross_prime
Structural signature: A networked or path-dependent flow system where the currently used path has become unsafe, unavailable, overloaded, or strategically suboptimal, and a viable alternate path exists or can be created.
Failure modes: diversion_to_bad_path, overload_migration, hidden_downstream_bottleneck, route_flapping, looped_routing, path_affinity_violation, stale_path_information, harm_displacement, unclear_destination_preservation, excessive_detour_cost, routing_black_hole, bypassing_necessary_controls, permanent_temporary_detour
Domain examples: networking_and_internet_routing, road_and_transportation_systems, supply_chain_and_logistics, software_service_routing, stormwater_and_hydraulic_systems, emergency_evacuation, organizational_escalation_paths, payment_and_financial_rails, public_service_intake_routing

Intent¶

Flow Diversion / Rerouting preserves continuity, safety, or viability by redirecting flow away from a blocked, overloaded, harmful, unavailable, costly, or strategically undesirable path toward an alternate viable path.

The archetype is useful when the flow itself should continue, but the current path has become a problem. Instead of stopping the flow, shedding the load, buffering it indefinitely, or switching the whole system into failover, the intervention changes the route.

In compact form:

When a flow path becomes blocked, harmful, overloaded, unavailable, or strategically undesirable, redirect the flow through an alternate viable path to preserve continuity at the cost of detour overhead and path-management complexity.

Primes¶

Composed of: Path Selection, Routing Rule, Alternate Path, Boundary, Observability, Constraint, Flow Control, Feedback, Path Health Signal

Structural Signature¶

This archetype is a strong candidate when the following conditions co-occur:

A flow moves through a path, channel, route, interface, dependency, corridor, supply line, escalation path, network edge, or process.
The current path is blocked, overloaded, hazardous, compromised, costly, unavailable, fragile, or otherwise undesirable.
The flow still has a purpose: reaching a destination, preserving function, avoiding harm, maintaining continuity, or entering a safe holding state.
At least one alternate path exists or can be created.
The system can detect the problematic path condition.
The system can redirect or reassign flow without losing critical state, meaning, authority, safety, or integrity.
The new path can be monitored for secondary overload or harm.

Flow Diversion / Rerouting is especially relevant when the failure is path-specific rather than demand-wide.

Intervention Signature¶

Detect path impairment or undesirability, select an alternate route, redirect flow through that route, and monitor whether the diversion preserves continuity without shifting harm or overload elsewhere.

The intervention changes the topology of movement:

source -> problematic path -> destination

to:

source -> alternate path -> destination

or, in more complex systems:

source -> route selection -> viable alternate path(s) -> preserved function

The key move is path substitution or path redirection while preserving the intended function of the flow.

Causal Logic¶

A flow system can fail because of the path it uses rather than because the flow itself is invalid or total demand is impossible. A network link congests. A road closes. A supplier fails. A drainage channel overflows. A customer request enters the wrong escalation path. A payment rail is unavailable. A service dependency becomes unsafe. A public-service intake route is blocked while another route has capacity.

Flow Diversion / Rerouting works by changing the path of propagation.

Path impairment becomes visible. The system detects blockage, overload, hazard, unavailability, cost, compromise, or undesirability.
Alternate routes are identified. The system determines where the flow can go instead.
Flow is redirected. The routing rule, operational process, or actor behavior changes the path.
The intended function is preserved. The flow still reaches a destination, service, safe state, or viable continuation.
Secondary effects are monitored. The alternate path is watched for overload, harm displacement, or unintended bottlenecks.
Reentry or revision remains possible. The system may return to the original path or select another route as conditions change.

The archetype transforms path-specific breakdown into adaptive path selection.

What It Is Not¶

Flow Diversion / Rerouting is not Load Balancing. Load Balancing distributes flow across multiple viable receivers or paths to improve utilization and reduce localized overload. Flow Diversion / Rerouting redirects flow away from a problematic path. Load balancing may include rerouting, but rerouting can be partial, emergency, tactical, or hazard-driven without balancing across all capacity.

Flow Diversion / Rerouting is not Failover. Failover transfers protected function to an alternate path or capacity after primary failure or degradation. Rerouting can occur before total failure, can be partial, and often changes the path of flow rather than ownership of function.

Flow Diversion / Rerouting is not Circuit Breaker. Circuit Breaker interrupts or meters flow at a boundary under overload or cascade risk. Rerouting preserves flow by changing where it goes.

Flow Diversion / Rerouting is not Backpressure. Backpressure slows or reshapes upstream production through downstream capacity signals. Rerouting changes the path, not necessarily the production rate.

Flow Diversion / Rerouting is not Buffering. Buffering holds flow temporarily. Rerouting sends flow elsewhere.

Flow Diversion / Rerouting is not Load Shedding. Load Shedding sacrifices load. Rerouting tries to preserve the flow by using another path.

Flow Diversion / Rerouting is not Decoupling via Interface. Interface decoupling stabilizes interaction by mediating dependency exposure. Rerouting changes the route through which flow travels.

Flow Diversion / Rerouting is not arbitrary redirection. A mature rerouting intervention preserves destination, function, safety, or meaning rather than simply moving flow out of sight.

Composition¶

Flow Diversion / Rerouting is composed from several lower-level abstractions:

Flow — Something must move through the system.
Network / topology — There must be distinguishable paths, routes, channels, nodes, or corridors.
Observability — Path impairment, congestion, hazard, or unavailability must be detectable.
Path selection — The system must choose among possible routes.
Alternate path — A viable substitute route must exist or be created.
Routing rule — Some rule, operator, protocol, or policy must redirect the flow.
Boundary — Rerouting often occurs at a decision point, junction, gate, interface, or admission boundary.
Feedback — Outcomes from rerouted flow should influence future path choice.
Constraint — Path choice must respect capacity, safety, cost, legality, state, or priority constraints.

The composition matters. Without an alternate path, diversion becomes stoppage. Without observability, rerouting is blind. Without path suitability checks, diversion may move harm elsewhere. Without feedback, rerouting may keep using a path after it becomes overloaded.

Mechanism Families¶

Common mechanism families include:

Network route-around failure or congestion — Traffic is routed through alternate links, paths, regions, or networks.
Traffic detours — Vehicles, pedestrians, or transit flows are redirected around closures, hazards, or congestion.
Supply-chain rerouting — Goods, materials, or orders are redirected through alternate suppliers, warehouses, carriers, or ports.
Logistics reassignment — Shipments or service obligations are routed through different fulfillment paths.
Drainage or hydraulic diversion — Water is redirected away from overloaded, blocked, or vulnerable channels.
Evacuation route planning — People are routed away from danger through alternate corridors.
Service request rerouting — Requests are sent to another team, queue, region, or provider when the ordinary path is blocked or unsuitable.
Organizational escalation rerouting — Decisions or issues bypass a blocked escalation chain and move through an alternate authority path.
Attention or demand redirection — Demand is redirected toward safer, less congested, or more appropriate channels.
Financial or payment rail rerouting — Transactions move through alternate payment channels when a rail is unavailable or risky.
Infrastructure bypass channels — Flow is routed around a damaged or constrained component through a bypass.

These mechanisms differ by domain, but they preserve the same intervention logic: change the path of flow to avoid a path-specific problem.

Parameter Dimensions¶

Concrete mechanisms usually require tuning along dimensions such as:

Diversion trigger threshold — What blockage, latency, hazard, cost, failure, or congestion level causes rerouting?
Alternate path selection rule — Which path is chosen and why?
Diversion fraction — How much flow is redirected?
Path capacity requirement — How much spare capacity must an alternate path have?
Path health signal weight — How strongly do health signals influence route choice?
Routing granularity — Is routing per item, request, batch, session, route, region, or time window?
Detour cost tolerance — How much extra cost is acceptable?
Maximum added latency or distance — How far or slow may the detour be?
Reentry condition — When can the original path be used again?
Rollback condition — When should the diversion be abandoned?
Path priority order — Which alternates are preferred?
Monitoring cadence — How often are path conditions reassessed?
Stakeholder notification rule — Who needs to know that flow has been diverted?
Load spread across alternates — Should diverted flow go to one alternate or several?

These are parameter dimensions, not the archetype itself.

Invariants to Preserve¶

Flow Diversion / Rerouting should preserve explicit invariants:

Critical function or destination is preserved — Rerouted flow should still serve its intended purpose.
Alternate path viability is checked — Flow should not be sent into an unsafe or incapable path.
No silent loss or corruption — Diverted flow should not disappear, duplicate incorrectly, or lose essential meaning.
No unbounded overload elsewhere — The alternate path should not collapse under diverted load.
Routing decisions are observable or auditable — Operators should know where flow went and why.
Safety and integrity bounds are preserved — The detour must not bypass necessary controls.
Reentry or route revision remains possible — The system should be able to change routes again as conditions evolve.
Affected parties are informed when needed — Stakeholders should not be misled when path change matters.

If these invariants cannot be preserved, it may be safer to stop, shed, buffer, or fail over rather than reroute.

Tradeoffs¶

Flow Diversion / Rerouting accepts path-management costs to preserve continuity.

Typical tradeoffs include:

Detour cost increases because alternate paths may be longer, slower, more expensive, or less efficient.
Latency or distance may increase because the best normal path is unavailable.
Alternate paths may become overloaded if diversion is too large.
Coordination overhead rises because path choice must be managed.
Routing complexity increases because path state must be monitored.
Path-specific optimization may be lost because the alternate route may not fit the flow as well.
Harm may be displaced from one path, region, group, or receiver to another.
Observability burden increases because downstream effects must be tracked.
Users or stakeholders may be confused if path changes are visible or disruptive.

The archetype is therefore not simply “go around.” It is a controlled topological adaptation.

Contraindications¶

Flow Diversion / Rerouting is a poor fit when no alternate path can safely preserve the flow's purpose.

Use cautiously or avoid when:

no viable alternate path exists,
the alternate path has lower safety or capacity than the problem path,
the destination or function cannot be preserved by rerouting,
the flow is path-dependent and cannot move without losing meaning, state, authority, or context,
path impairment signals are unreliable,
diversion would shift unacceptable harm to others,
routing changes would violate authority, legal, contractual, or safety constraints,
total demand exceeds all available paths and requires load shedding or rate limiting,
path change would break state affinity, ownership, or context.

In such cases, failover, circuit breaking, buffering, backpressure, load shedding, capacity expansion, repair, or redesign may be more appropriate.

Failure Modes¶

Common failure modes include:

Diversion to bad path — Flow is sent to an alternate path that cannot safely handle it.
Overload migration — The bottleneck moves rather than disappearing.
Hidden downstream bottleneck — The alternate route appears open but shares a constrained downstream dependency.
Route flapping — Flow repeatedly switches between paths as conditions fluctuate.
Looped routing — Flow circulates without reaching destination.
Path affinity violation — Flow loses necessary state, authority, ownership, or context.
Stale path information — Decisions are based on outdated route conditions.
Harm displacement — The intervention protects one area by burdening another unfairly or dangerously.
Unclear destination preservation — The diverted flow no longer fulfills the original purpose.
Excessive detour cost — Continuity is preserved at an unacceptable cost.
Routing black hole — Flow is redirected into a path where it disappears or stalls.
Bypassing necessary controls — The detour avoids safety, compliance, validation, or review gates.
Permanent temporary detour — A temporary reroute becomes the unreviewed default.

These failure modes should be treated as part of the archetype's design space.

Worked Example¶

A logistics company normally routes shipments through a major port because it is fast, cheap, and well integrated with downstream warehouses. A labor disruption and severe congestion make the port unreliable. Containers begin sitting for days, customer deadlines are missed, and the backlog grows.

The company implements Flow Diversion / Rerouting.

Congestion and dwell-time thresholds define when the ordinary port becomes undesirable.
Shipments are classified by deadline, destination, and handling requirements.
Some shipments are rerouted through a secondary port.
Others are redirected to rail transfer points or alternate warehouses.
The company monitors added cost, delivery time, and congestion at the alternate routes.
When the original port stabilizes, the company gradually returns appropriate flows.

The intervention does not reduce demand. It does not simply fail over the whole system to a single backup. It changes the paths used by different flows so the shipment function can continue despite path-specific disruption.

The key move is rerouting flow around the impaired path while preserving delivery function.

Cross-Domain Instances¶

Networking and internet routing — Packets or traffic are routed around failed, congested, or high-cost paths.
Road and transportation systems — Vehicles, pedestrians, or transit flows are diverted around closures, hazards, or congestion.
Supply chain and logistics — Goods are rerouted through alternate suppliers, warehouses, ports, carriers, or corridors.
Software service routing — Requests are redirected to alternate regions, services, endpoints, or queues when a path is impaired.
Stormwater and hydraulic systems — Water is diverted away from overloaded channels or vulnerable areas.
Emergency evacuation — People are routed away from danger through viable evacuation corridors.
Organizational escalation paths — Decisions or issues are rerouted around blocked, conflicted, overloaded, or unavailable authorities.
Payment and financial rails — Transactions are routed through alternate rails when one channel is unavailable, risky, or too slow.
Public service intake routing — Requests or cases are redirected to less congested offices, regions, or service channels.

These examples are structurally related because each redirects flow away from a problematic path toward an alternate route while preserving the intended function.

Notes¶

Flow Diversion / Rerouting should be reviewed alongside Load Balancing, Failover, Circuit Breaker, Backpressure, Buffering, Rate Limiting, Load Shedding, Decoupling via Interface, and Bulkhead Isolation.

The main conceptual risk is collapse into nearby concepts:

If the entry emphasizes distributing live load across multiple viable receivers, it becomes Load Balancing.
If the entry emphasizes transferring protected function after primary failure, it becomes Failover.
If the entry emphasizes interrupting flow under cascade risk, it becomes Circuit Breaker.
If the entry emphasizes holding flow temporarily, it becomes Buffering.
If the entry emphasizes reducing admitted flow, it becomes Rate Limiting.
If the entry emphasizes sacrificing flow, it becomes Load Shedding.
If the entry emphasizes stabilizing interaction across a boundary, it becomes Decoupling via Interface.

The current entry uses path_selection, routing_rule, alternate_path, flow_control, and path_health_signal as solution-side labels. These may need later normalization as lower-level archetypal components, prime abstractions, mechanisms, or informal component labels.