Spanda

Recovery Simulation

The Recovery Orchestrator supports failure simulation without affecting production systems.

Modes

Mode	CLI	API	Side effects
Plan	`spanda recovery plan`	`POST /v1/recovery/plan`	None — generates plans only
Simulate	`spanda recovery simulate`	`POST /v1/recovery/simulate`	None — timelines + mission impact
Dry-run	`spanda recovery dry-run`	`POST /v1/recovery/validate`	None — full pipeline, no execution
Validate	`spanda recovery validate`	`POST /v1/recovery/validate`	Runs validation gates
Execute	`spanda recovery execute --force`	`POST /v1/recovery/execute`	Delegates to assurance execution

Examples

# Plan recovery for GPS loss on a robot
spanda recovery plan rover.sd --entity robot-1 --failure gps_loss --json

# Simulate sensor failure with expected downtime
spanda recovery simulate rover.sd --failure sensor_failure

# Dry-run full recovery pipeline
spanda recovery dry-run rover.sd --entity robot-1 --failure connectivity_loss

# Validate gates without execution
spanda recovery validate rover.sd --failure lidar.failed

Output

Simulation reports include:

Orchestrated recovery plans with strategies and escalation levels
Upstream/downstream impact from the recovery graph
Decision explanations (can/should/safe/authorized)
Estimated duration and mission disruption score
Predictive indicators (when telemetry provided) — also via GET /v1/recovery/predictive
Strategy recommendations — POST /v1/recovery/recommend
Legacy assurance report for backward compatibility

Spanda

Recovery Simulation

Modes

Examples

Output

See also