In a Kappa architecture (stream processing only, no separate batch layer), how do you recompute results after fixing a bug?

Options

• ✓ Replay the retained log through the fixed job into a new output, then switch readers over Correct
• ✗ Run a separate batch job over the database
• ✗ Ask producers to resend their events
• ✗ It is not possible; results stay wrong

Why

The log is the source of truth, kept long enough to replay. Lambda architecture instead runs a batch layer beside the stream, which means writing the same logic twice.

Review this in your daily deck → All cards in Streaming