PATTERN Cited by 1 source

Fallback on New Connector Failure

Definition

During a system migration, a conditional workflow step guarantees that if the new implementation fails, the workflow immediately re-executes using the legacy implementation — ensuring zero data-loss and zero user-visible failure. The user may observe slightly longer runtime but never a failed job.

Canonical example

Netflix's Cassandra data-movement migration (Source: sources/2026-06-19-netflix-the-evolution-of-cassandra-data-movement-at-netflix):

• Maestro workflow includes a conditional step: if the Move Data connector step fails → execute the Casspactor step.
• Users never see a migration failure or stale data — at worst, runtime increases slightly.

When to apply

• Migrating a mission-critical pipeline where failure means stale data or SLA breach.
• The legacy system is still operational and can serve as a safety net during rollout.
• Combined with patterns/decider-pattern-workflow-routing for per-job routing control.

Trade-offs

• Pro: Eliminates blast radius during migration; builds confidence in new system incrementally.
• Con: Doubles infrastructure cost during migration window (both systems must remain operational); legacy degradation must be monitored.

Seen in

• sources/2026-06-19-netflix-the-evolution-of-cassandra-data-movement-at-netflix — safety pillar of Netflix's three-pillar migration strategy.