org.apache.flink.runtime.webmonitor.threadinfo

类 ThreadInfoRequestCoordinator

java.lang.Object

org.apache.flink.runtime.webmonitor.stats.TaskStatsRequestCoordinator<Map<ExecutionAttemptID,Collection<ThreadInfoSample>>,VertexThreadInfoStats>

org.apache.flink.runtime.webmonitor.threadinfo.ThreadInfoRequestCoordinator

public class ThreadInfoRequestCoordinator
extends TaskStatsRequestCoordinator<Map<ExecutionAttemptID,Collection<ThreadInfoSample>>,VertexThreadInfoStats>

A coordinator for triggering and collecting thread info stats of running job vertex subtasks.

嵌套类概要

从类继承的嵌套类/接口 org.apache.flink.runtime.webmonitor.stats.TaskStatsRequestCoordinator

TaskStatsRequestCoordinator.PendingStatsRequest<T,V>

字段概要

从类继承的字段 org.apache.flink.runtime.webmonitor.stats.TaskStatsRequestCoordinator

executor, isShutDown, lock, log, NUM_GHOST_SAMPLE_IDS, pendingRequests, recentPendingRequestIds, requestIdCounter, requestTimeout

构造器概要

构造器

构造器和说明
ThreadInfoRequestCoordinator(Executor executor, java.time.Duration requestTimeout) Creates a new coordinator for the job.

方法概要

限定符和类型	方法和说明
CompletableFuture<VertexThreadInfoStats>	triggerThreadInfoRequest(Map<org.apache.flink.shaded.guava30.com.google.common.collect.ImmutableSet<ExecutionAttemptID>,CompletableFuture<TaskExecutorThreadInfoGateway>> executionsWithGateways, int numSamples, java.time.Duration delayBetweenSamples, int maxStackTraceDepth) Triggers collection of thread info stats of a job vertex by combining thread info responses from given subtasks.

从类继承的方法 org.apache.flink.runtime.webmonitor.stats.TaskStatsRequestCoordinator

getNumberOfPendingRequests, handleFailedResponse, handleSuccessfulResponse, shutDown

从类继承的方法 java.lang.Object

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

构造器详细资料

ThreadInfoRequestCoordinator

public ThreadInfoRequestCoordinator(Executor executor,
                                    java.time.Duration requestTimeout)

Creates a new coordinator for the job.

参数:

executor - Used to execute the futures.

requestTimeout - Time out after the expected sampling duration. This is added to the expected duration of a request, which is determined by the number of samples and the delay between each sample.

方法详细资料

triggerThreadInfoRequest

public CompletableFuture<VertexThreadInfoStats> triggerThreadInfoRequest(Map<org.apache.flink.shaded.guava30.com.google.common.collect.ImmutableSet<ExecutionAttemptID>,CompletableFuture<TaskExecutorThreadInfoGateway>> executionsWithGateways,
                                                                         int numSamples,
                                                                         java.time.Duration delayBetweenSamples,
                                                                         int maxStackTraceDepth)

Triggers collection of thread info stats of a job vertex by combining thread info responses from given subtasks. A thread info response of a subtask in turn consists of numSamples, collected with delayBetweenSamples milliseconds delay between them.

参数:

executionsWithGateways - Execution attempts together with TaskExecutors running them.

numSamples - Number of thread info samples to collect from each subtask.

delayBetweenSamples - Delay between consecutive samples (ms).

maxStackTraceDepth - Maximum depth of the stack traces collected within thread info samples.

返回:

A future of the completed thread info stats.