FailureIntervalSweepScenario
Reliability scenarios · Implementation: mlsysim/engine/scenario_definitions/reliability.py.
Construct this scenario with keyword inputs. The fields below show the types and defaults declared by this class.
Interval between failures of component_mtbf components, at each job size in sizes.
Notes
Equation: Compose mlsysim.physics.reliability.calc_mtbf_cluster.
Assumptions: Components have identical independent stationary failure hazards; their hazards add with fleet size.
Excludes: Correlated failures, repairs, and time-varying failure rates are absent.
References
Derivation: this scenario binds its input fields to the referenced implementation contracts: mlsysim.physics.reliability.calc_mtbf_cluster. The methods below identify any additional arithmetic.
Examples
>>> from mlsysim import Scenarios
>>> case = Scenarios.GpuFailureLadder
>>> str(next(iter(case.intervals.values())))
'6250.0 h'Attributes
| Name | Description |
|---|---|
| component_mtbf | |
| failures_per_day | Expected failure counts per day at each job size. Equation: one day divided by each mean |
| intervals | Mean failure intervals indexed by component count. Equation: |
| sizes |
component_mtbf
Type: Quantity.
Declared default: required.
sizes
Type: List[int].
Declared default: required.
intervals
engine.scenario_types.FailureIntervalSweepScenario.intervals
Mean failure intervals indexed by component count. Equation: mlsysim.physics.reliability.calc_mtbf_cluster(component_mtbf, n) for each size. Units: time; independent identical failure hazards are assumed.
failures_per_day
engine.scenario_types.FailureIntervalSweepScenario.failures_per_day
Expected failure counts per day at each job size. Equation: one day divided by each mean interval. Units: dimensionless counts per stated day, not event probabilities.