What is retry.backoff.ms?

• A. The maximum time to wait for a broker response.
• B. The amount of time to wait before attempting to retry a failed request to a given topic partition.
• C. The time to wait before rebalancing partitions.
• D. The timeout for consumer group coordination

Answer: (B)

The idea being tested is how the client handles retries after a failure. retry.backoff.ms is the pause you wait after a failed request before trying again. When a producer (or client) sends data to a broker and the request fails due to a transient issue, the client doesn’t immediately retry. It waits retry.backoff.ms milliseconds and then retries the request to the same topic partition. This backoff helps prevent a flood of retries and gives the broker or network time to recover or refresh metadata.

This setting is not about the total time allowed for a broker response (that’s governed by a separate timeout) and it’s not related to partition rebalancing or consumer group coordination timeouts.