[KAFKA-1211] Hold the produce request with ack > 1 in purgatory until replicas' HW has larger than the produce offset (KIP-101)

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 0.11.0.0
• Labels: reliability
• Assignee: Ben Stopford
• Reporter: Guozhang Wang
• Created: 16/Jan/14 21:13
• Updated: 07/Apr/17 14:33
• Resolved: 06/Apr/17 21:53

Description

Today during leader failover we will have a weakness period when the followers truncate their data before fetching from the new leader, i.e., number of in-sync replicas is just 1. If during this time the leader has also failed then produce requests with ack >1 that have get responded will still be lost. To avoid this scenario we would prefer to hold the produce request in purgatory until replica's HW has larger than the offset instead of just their end-of-log offsets.

Issue Links

• is blocked by: KAFKA-1188 Stale LeaderAndIsr request could be handled by the broker on Controller failover
• is related to: KAFKA-3919 Broker faills to start after ungraceful shutdown due to non-monotonically incrementing offsets in logs
• mentioned in: Page Loading...