Package org.apache.pinot.query.runtime.plan.pipeline

Class PipelineBreakerExecutor

java.lang.Object

org.apache.pinot.query.runtime.plan.pipeline.PipelineBreakerExecutor

public class PipelineBreakerExecutor
extends Object

Utility class to run pipeline breaker execution and collects the results.

Method Summary

Modifier and Type	Method	Description
static PipelineBreakerResult	executePipelineBreakers(OpChainSchedulerService scheduler, MailboxService mailboxService, DistributedStagePlan distributedStagePlan, Map<String,String> requestMetadata, long requestId, long deadlineMs)	Execute a pipeline breaker and collect the results (synchronously).

Methods inherited from class java.lang.Object

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

Method Detail

executePipelineBreakers

@Nullable
public static PipelineBreakerResult executePipelineBreakers(OpChainSchedulerService scheduler,
                                                            MailboxService mailboxService,
                                                            DistributedStagePlan distributedStagePlan,
                                                            Map<String,String> requestMetadata,
                                                            long requestId,
                                                            long deadlineMs)

Execute a pipeline breaker and collect the results (synchronously). Currently, pipeline breaker executor can only execute mailbox receive pipeline breaker.

Parameters:

scheduler - scheduler service to run the pipeline breaker main thread.

mailboxService - mailbox service to attach the MailboxReceiveNode against.

distributedStagePlan - the distributed stage plan to run pipeline breaker on.

requestMetadata - request metadata, including query options

requestId - request ID

deadlineMs - execution deadline

Returns:

pipeline breaker result; - If exception occurs, exception block will be wrapped in TransferableBlock and assigned to each PB node. - Normal stats will be attached to each PB node and downstream execution should return with stats attached.