CONCEPT Cited by 1 source
SLA-based eventual consistency¶
Definition¶
SLA-based eventual consistency is a consistency model where a system guarantees convergence to the correct state within a predictable, operationally bounded time window (the SLA), rather than requiring immediate consistency or leaving convergence unbounded. The key distinction from generic eventual consistency: the convergence timeline is an explicit operational contract, not a best-effort aspiration.
When it's acceptable¶
SLA-based eventual consistency works when:
- Downstream consumers do not need real-time reflection of upstream changes
- The system can tolerate temporary divergence (e.g., null values for new columns)
- The bounded window is short enough that operational risk remains manageable
- Deployment sequencing can restore full consistency predictably
Pinterest's instance¶
Pinterest applies SLA-based eventual consistency to schema evolution in their CDC pipeline:
"An SLA-based model is acceptable at Pinterest because schema changes are not usually consumed in real time — what matters is that the new schema becomes available within a predictable, operationally safe window." (Source: sources/2026-06-24-pinterest-automated-schema-evolution-in-pinterests-next-generation-db)
The bounded window spans three phases (schema → code → data convergence), with deployment sequencing (Spark first for backfill, Flink second for new data) ensuring full consistency is restored within the SLA.
Seen in¶
- sources/2026-06-24-pinterest-automated-schema-evolution-in-pinterests-next-generation-db — canonical instance: Pinterest's three-phase schema convergence with bounded divergence window