Package org.apache.druid.server.coordinator.loading

Class ReplicationThrottler

java.lang.Object

org.apache.druid.server.coordinator.loading.ReplicationThrottler

public class ReplicationThrottler
extends Object

The ReplicationThrottler is used to throttle the number of segment replicas that are assigned to a load queue in a single run. This is achieved by setting the values of the following configs:

• CoordinatorDynamicConfig.getReplicationThrottleLimit() - Maximum number of replicas that can be assigned to a tier in a single run.

Constructor Summary

Constructors

Constructor	Description
ReplicationThrottler(Map<String,Integer> tierToLoadingReplicaCount, int replicationThrottleLimit)	Creates a new ReplicationThrottler for use during a single coordinator run.

Method Summary

Modifier and Type	Method	Description
void	incrementAssignedReplicas(String tier)
boolean	isReplicationThrottledForTier(String tier)

Methods inherited from class java.lang.Object

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

Constructor Details

ReplicationThrottler

public ReplicationThrottler(Map<String,Integer> tierToLoadingReplicaCount,
 int replicationThrottleLimit)

Creates a new ReplicationThrottler for use during a single coordinator run. The number of replicas loading on a tier must always be within the current replicationThrottleLimit. Thus, if a tier was already loading k replicas at the start of a coordinator run, it may be assigned only replicationThrottleLimit - k more replicas during the run.

Parameters:

tierToLoadingReplicaCount - Map from tier name to number of replicas already being loaded.

replicationThrottleLimit - Maximum number of replicas that can be assigned to a single tier in the current run.

Method Details

isReplicationThrottledForTier

public boolean isReplicationThrottledForTier(String tier)

incrementAssignedReplicas

public void incrementAssignedReplicas(String tier)