Retry failed batches in large data syncs

Large data syncs can now retry individual batches that fail. When retry is enabled on a batched flow, Prismatic re-runs only the failed batch, with the same records, as part of the same execution. The sync's other batches keep processing in the meantime, so one transient error no longer costs you the rest of a long sync.

Batch retry is opt-in and uses the flow's existing retry settings:

• Low-code: Select your trigger, open its Flow retry tab, and toggle Retry.
• Code-native: Add a retryConfig to a flow that uses batchFlowTrigger.

Batched flows support up to 3 retry attempts and up to 15 minutes between attempts, with optional exponential backoff. If a batched flow already has retry enabled, its failed batches are now retried individually, and the whole execution is no longer retried.

For details, see Retrying failed batches in large data syncs and Large Data Sync.