Operational primitive
A system-agnostic unit of distributed system behavior that forms the basic vocabulary of the pattern library. Examples: BACKPRESSURE_ACCUMULATION, QUEUE_GROWTH, TAIL_LATENCY_EXPANSION. Primitives describe behavioral dynamics rather than vendor metrics, making learned patterns portable across infrastructure environments.
Operational primitives are the vocabulary layer that makes the pattern library portable. Infrastructure environments differ enormously in their metric names and tooling: one environment emits kafka_consumer_lag_sum, another emits messaging.consumer.lag, another has it buried in a proprietary dashboard. But both are expressing the same behavioral dynamic: a consumer falling behind its producer. By mapping both vendor-specific metrics to the primitive BACKPRESSURE_ACCUMULATION via a system mapping, the pattern library can match the same behavioral signature regardless of which infrastructure produced it. The set of defined primitives covers the fundamental dynamics of distributed systems: queue dynamics (QUEUE_GROWTH, QUEUE_DEPTH_SPIKE), latency patterns (TAIL_LATENCY_EXPANSION, P99_DIVERGENCE), resource pressures (MEMORY_PRESSURE, CPU_SATURATION, DISK_IO_CONTENTION), and reliability signals (ERROR_RATE_ELEVATION, RETRY_STORM, CASCADE_FAILURE_PRECURSOR). Patterns are defined as co-occurring sets of these primitives rather than co-occurring vendor metrics, which is why a pattern learned in one Kubernetes cluster immediately applies to a new cluster running different monitoring software.