Batch Processing

Prove ordering across repeated batches in a sequencer loop.

Batch processing typically runs continuously:

  1. collect transactions for a time window
  2. order them deterministically
  3. generate a proof
  4. publish batch + proof reference

The main production considerations are:

  • stable ordering inputs (hash + ingress time)
  • robust persistence of proof metadata
  • backpressure when proof generation latency increases