kafka.server

Class ReplicaFetcherTierStateMachine

java.lang.Object

kafka.server.ReplicaFetcherTierStateMachine

All Implemented Interfaces:

TierStateMachine

public class ReplicaFetcherTierStateMachine
extends Object
implements TierStateMachine

The replica fetcher tier state machine follows a state machine progression. Currently, the tier state machine follows a synchronous execution, and we only need to start the machine. There is no need to advance the state. When started, the tier state machine will fetch the local log start offset of the leader and then build the follower's remote log aux state until the leader's local log start offset.

Constructor Summary

Constructors

Constructor and Description
ReplicaFetcherTierStateMachine(kafka.server.LeaderEndPoint leader, kafka.server.ReplicaManager replicaMgr)

Method Summary

Modifier and Type	Method and Description
Optional<kafka.server.PartitionFetchState>	maybeAdvanceState(org.apache.kafka.common.TopicPartition topicPartition, kafka.server.PartitionFetchState currentFetchState) This is currently a no-op but will be used for implementing async tiering logic in KAFKA-13560.
kafka.server.PartitionFetchState	start(org.apache.kafka.common.TopicPartition topicPartition, kafka.server.PartitionFetchState currentFetchState, org.apache.kafka.common.message.FetchResponseData.PartitionData fetchPartitionData) Start the tier state machine for the provided topic partition.

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

ReplicaFetcherTierStateMachine

public ReplicaFetcherTierStateMachine(kafka.server.LeaderEndPoint leader,
                                      kafka.server.ReplicaManager replicaMgr)

Method Detail

start

public kafka.server.PartitionFetchState start(org.apache.kafka.common.TopicPartition topicPartition,
                                              kafka.server.PartitionFetchState currentFetchState,
                                              org.apache.kafka.common.message.FetchResponseData.PartitionData fetchPartitionData)
                                       throws Exception

Start the tier state machine for the provided topic partition. Currently, this start method will build the entire remote aux log state synchronously.

Specified by:

start in interface TierStateMachine

Parameters:

topicPartition - the topic partition

currentFetchState - the current PartitionFetchState which will be used to derive the return value

fetchPartitionData - the data from the fetch response that returned the offset moved to tiered storage error

Returns:

the new PartitionFetchState after the successful start of the tier state machine

Throws:

Exception

maybeAdvanceState

public Optional<kafka.server.PartitionFetchState> maybeAdvanceState(org.apache.kafka.common.TopicPartition topicPartition,
                                                                    kafka.server.PartitionFetchState currentFetchState)

This is currently a no-op but will be used for implementing async tiering logic in KAFKA-13560.

Specified by:

maybeAdvanceState in interface TierStateMachine

Parameters:

topicPartition - the topic partition

currentFetchState - the current PartitionFetchState which will be used to derive the return value

Returns:

the original PartitionFetchState