DAGScheduler — Stage-Oriented Scheduler

Note

The introduction that follows was highly influenced by the scaladoc of org.apache.spark.scheduler.DAGScheduler. As DAGScheduler is a private class it does not appear in the official API documentation. You are strongly encouraged to read the sources and only then read this and the related pages afterwards.

"Reading the sources", I say?! Yes, I am kidding!

Introduction

DAGScheduler is the scheduling layer of Apache Spark that implements stage-oriented scheduling. It transforms a logical execution plan (i.e. RDD lineage of dependencies built using RDD transformations) to a physical execution plan (using stages).

dagscheduler rdd lineage stage dag.png
Figure 1. DAGScheduler Transforming RDD Lineage Into Stage DAG

After an action has been called, SparkContext hands over a logical plan to DAGScheduler that it in turn translates to a set of stages that are submitted as TaskSets for execution (see Execution Model).

dagscheduler rdd partitions job resultstage.png
Figure 2. Executing action leads to new ResultStage and ActiveJob in DAGScheduler

The fundamental concepts of DAGScheduler are jobs and stages (refer to Jobs and Stages respectively) that it tracks through internal registries and counters.

DAGScheduler works solely on the driver and is created as part of SparkContext’s initialization (right after TaskScheduler and SchedulerBackend are ready).

dagscheduler new instance.png
Figure 3. DAGScheduler as created by SparkContext with other services

DAGScheduler does three things in Spark (thorough explanations follow):

  • Computes an execution DAG, i.e. DAG of stages, for a job.

  • Determines the preferred locations to run each task on.

  • Handles failures due to shuffle output files being lost.

DAGScheduler computes a directed acyclic graph (DAG) of stages for each job, keeps track of which RDDs and stage outputs are materialized, and finds a minimal schedule to run jobs. It then submits stages to TaskScheduler.

In addition to coming up with the execution DAG, DAGScheduler also determines the preferred locations to run each task on, based on the current cache status, and passes the information to TaskScheduler.

DAGScheduler tracks which RDDs are cached (or persisted) to avoid "recomputing" them, i.e. redoing the map side of a shuffle. DAGScheduler remembers what ShuffleMapStages have already produced output files (that are stored in BlockManagers).

DAGScheduler is only interested in cache location coordinates, i.e. host and executor id, per partition of a RDD.

Caution
FIXME: A diagram, please

Furthermore, it handles failures due to shuffle output files being lost, in which case old stages may need to be resubmitted. Failures within a stage that are not caused by shuffle file loss are handled by the TaskScheduler itself, which will retry each task a small number of times before cancelling the whole stage.

DAGScheduler uses an event queue architecture in which a thread can post DAGSchedulerEvent events, e.g. a new job or stage being submitted, that DAGScheduler reads and executes sequentially. See the section Internal Event Loop - dag-scheduler-event-loop.

DAGScheduler runs stages in topological order.

Table 1. DAGScheduler Internal Registries and Counters
Name Description

activeJobs

ActiveJob instances

cacheLocs

Block locations per RDD and partition.

Uses TaskLocation that includes a host name and an executor id on that host (as ExecutorCacheTaskLocation).

The keys are RDDs (their ids) and the values are arrays indexed by partition numbers.

Each entry is a set of block locations where a RDD partition is cached, i.e. the BlockManagers of the blocks.

Initialized empty when DAGScheduler is created.

Used when DAGScheduler is requested for the locations of the cache blocks of a RDD or clear them.

failedEpoch

The lookup table of lost executors and the epoch of the event.

failedStages

Stages that failed due to fetch failures (when a task fails with FetchFailed exception).

jobIdToActiveJob

The lookup table of ActiveJobs per job id.

jobIdToStageIds

The lookup table of all stages per ActiveJob id

nextJobId

The next job id counting from 0.

Used when DAGScheduler submits a job and a map stage, and runs an approximate job.

nextStageId

The next stage id counting from 0.

Used when DAGScheduler creates a shuffle map stage and a result stage. It is the key in stageIdToStage.

runningStages

The set of stages that are currently "running".

A stage is added when submitMissingTasks gets executed (without first checking if the stage has not already been added).

shuffleIdToMapStage

The lookup table of ShuffleMapStages per ShuffleDependency.

stageIdToStage

The lookup table for stages per their ids.

Used when DAGScheduler creates a shuffle map stage, creates a result stage, cleans up job state and independent stages, is informed that a task is started, a taskset has failed, a job is submitted (to compute a ResultStage), a map stage was submitted, a task has completed or a stage was cancelled, updates accumulators, aborts a stage and fails a job and independent stages.

waitingStages

The stages with parents to be computed

Tip

Enable INFO, DEBUG or TRACE logging levels for org.apache.spark.scheduler.DAGSchedule logger to see what happens inside DAGScheduler.

Add the following line to conf/log4j.properties:

log4j.logger.org.apache.spark.scheduler.DAGScheduler=TRACE

Refer to Logging.

DAGScheduler uses SparkContext, TaskScheduler, LiveListenerBus, MapOutputTracker and BlockManager for its services. However, at the very minimum, DAGScheduler takes a SparkContext only (and requests SparkContext for the other services).

DAGScheduler reports metrics about its execution (refer to the section Metrics).

When DAGScheduler schedules a job as a result of executing an action on a RDD or calling SparkContext.runJob() method directly, it spawns parallel tasks to compute (partial) results per partition.

Running Approximate Job — runApproximateJob Method

Caution
FIXME

createResultStage Internal Method

createResultStage(
  rdd: RDD[_],
  func: (TaskContext, Iterator[_]) => _,
  partitions: Array[Int],
  jobId: Int,
  callSite: CallSite): ResultStage
Caution
FIXME

updateJobIdStageIdMaps Method

Caution
FIXME

Creating DAGScheduler Instance

DAGScheduler takes the following when created:

DAGScheduler initializes the internal registries and counters.

DAGScheduler sets itself in the given TaskScheduler and in the end starts eventProcessLoop.

Note
DAGScheduler can reference all the services through a single SparkContext with or without specifying explicit TaskScheduler.

LiveListenerBus Event Bus for SparkListenerEvents — listenerBus Property

listenerBus: LiveListenerBus

listenerBus is a LiveListenerBus to post scheduling events and is passed in when DAGScheduler is created.

executorHeartbeatReceived Method

executorHeartbeatReceived(
  execId: String,
  accumUpdates: Array[(Long, Int, Int, Seq[AccumulableInfo])],
  blockManagerId: BlockManagerId): Boolean

executorHeartbeatReceived posts a SparkListenerExecutorMetricsUpdate (to listenerBus) and informs BlockManagerMaster that blockManagerId block manager is alive (by posting BlockManagerHeartbeat).

Note
executorHeartbeatReceived is called when TaskSchedulerImpl handles executorHeartbeatReceived.

Cleaning Up After ActiveJob and Independent Stages — cleanupStateForJobAndIndependentStages Method

cleanupStateForJobAndIndependentStages(job: ActiveJob): Unit

cleanupStateForJobAndIndependentStages cleans up the state for job and any stages that are not part of any other job.

cleanupStateForJobAndIndependentStages looks the job up in the internal jobIdToStageIds registry.

If no stages are found, the following ERROR is printed out to the logs:

ERROR No stages registered for job [jobId]

Oterwise, cleanupStateForJobAndIndependentStages uses stageIdToStage registry to find the stages (the real objects not ids!).

For each stage, cleanupStateForJobAndIndependentStages reads the jobs the stage belongs to.

If the job does not belong to the jobs of the stage, the following ERROR is printed out to the logs:

ERROR Job [jobId] not registered for stage [stageId] even though that stage was registered for the job

If the job was the only job for the stage, the stage (and the stage id) gets cleaned up from the registries, i.e. runningStages, shuffleIdToMapStage, waitingStages, failedStages and stageIdToStage.

While removing from runningStages, you should see the following DEBUG message in the logs:

DEBUG Removing running stage [stageId]

While removing from waitingStages, you should see the following DEBUG message in the logs:

DEBUG Removing stage [stageId] from waiting set.

While removing from failedStages, you should see the following DEBUG message in the logs:

DEBUG Removing stage [stageId] from failed set.

After all cleaning (using stageIdToStage as the source registry), if the stage belonged to the one and only job, you should see the following DEBUG message in the logs:

DEBUG After removal of stage [stageId], remaining stages = [stageIdToStage.size]

The job is removed from jobIdToStageIds, jobIdToActiveJob, activeJobs registries.

The final stage of the job is removed, i.e. ResultStage or ShuffleMapStage.

Marking ShuffleMapStage Job Finished — markMapStageJobAsFinished Method

markMapStageJobAsFinished(job: ActiveJob, stats: MapOutputStatistics): Unit

markMapStageJobAsFinished marks the active job finished and notifies Spark listeners.

Internally, markMapStageJobAsFinished marks the zeroth partition finished and increases the number of tasks finished in job.

Ultimately, SparkListenerJobEnd is posted to LiveListenerBus (as listenerBus) for the job, the current time (in millis) and JobSucceeded job result.

Note
markMapStageJobAsFinished is used in handleMapStageSubmitted and handleTaskCompletion.

Submitting Job — submitJob method

submitJob[T, U](
  rdd: RDD[T],
  func: (TaskContext, Iterator[T]) => U,
  partitions: Seq[Int],
  callSite: CallSite,
  resultHandler: (Int, U) => Unit,
  properties: Properties): JobWaiter[U]

submitJob creates a JobWaiter and posts a JobSubmitted event.

dagscheduler submitjob.png
Figure 4. DAGScheduler.submitJob

Internally, submitJob does the following:

  1. Checks whether partitions reference available partitions of the input rdd.

  2. Increments nextJobId internal job counter.

  3. Returns a 0-task JobWaiter when the number of partitions is zero.

  4. Posts a JobSubmitted event and returns a JobWaiter.

You may see a IllegalArgumentException thrown when the input partitions references partitions not in the input rdd:

Attempting to access a non-existent partition: [p]. Total number of partitions: [maxPartitions]
Note
submitJob is called when SparkContext submits a job and DAGScheduler runs a job.
Note
submitJob assumes that the partitions of a RDD are indexed from 0 onwards in sequential order.

Submitting ShuffleDependency for Execution — submitMapStage Method

submitMapStage[K, V, C](
  dependency: ShuffleDependency[K, V, C],
  callback: MapOutputStatistics => Unit,
  callSite: CallSite,
  properties: Properties): JobWaiter[MapOutputStatistics]

submitMapStage creates a JobWaiter (that it eventually returns) and posts a MapStageSubmitted event to DAGScheduler’s Internal Event Bus).

Internally, submitMapStage increments nextJobId internal counter to get the job id.

submitMapStage then creates a JobWaiter (with the job id and with one artificial task that will however get completed only when the entire stage finishes).

submitMapStage announces the map stage submission application-wide (by posting a MapStageSubmitted to LiveListenerBus).

Note
A MapStageSubmitted holds the newly-created job id and JobWaiter with the input dependency, callSite and properties parameters.

submitMapStage returns the JobWaiter.

If the number of partition to compute is 0, submitMapStage throws a SparkException:

Can't run submitMapStage on RDD with 0 partitions
Note
submitMapStage is used when SparkContext submits a map stage for execution.

Posting StageCancelled Event — cancelStage Method

cancelStage(stageId: Int)

cancelJobGroup merely posts a StageCancelled event to the DAGScheduler’s Internal Event Bus.

Note
cancelStage is executed when a SparkContext is requested to cancel a stage.

Posting JobGroupCancelled Event — cancelJobGroup Method

cancelJobGroup(groupId: String): Unit

cancelJobGroup prints the following INFO message to the logs followed by posting a JobGroupCancelled event to the DAGScheduler’s Internal Event Bus.

INFO Asked to cancel job group [groupId]
Note
cancelJobGroup is executed when a SparkContext is requested to cancel a specified group of jobs.

Posting AllJobsCancelled Event — cancelAllJobs Method

cancelAllJobs(): Unit

cancelAllJobs merely posts a AllJobsCancelled event to the DAGScheduler’s Internal Event Bus.

Posting BeginEvent Event — taskStarted Method

taskStarted(task: Task[_], taskInfo: TaskInfo)

taskStarted merely posts a BeginEvent event to the DAGScheduler’s Internal Event Bus.

Note
taskStarted is executed when a TaskSetManager starts a task.

Posting GettingResultEvent Event — taskGettingResult Method

taskGettingResult(taskInfo: TaskInfo)

taskGettingResult merely posts a GettingResultEvent event to the DAGScheduler’s Internal Event Bus.

Note
taskGettingResult is executed when a TaskSetManager gets notified about a task fetching result.

Announcing Task End (Posting CompletionEvent to DAGScheduler Event Bus) — taskEnded Method

taskEnded(
  task: Task[_],
  reason: TaskEndReason,
  result: Any,
  accumUpdates: Map[Long, Any],
  taskInfo: TaskInfo,
  taskMetrics: TaskMetrics): Unit

taskEnded simply posts a CompletionEvent event to the DAGScheduler’s Internal Event Bus.

Note
taskEnded is called when a TaskSetManager reports task completions, i.e. success or failure.
Tip
Read TaskMetrics.

Posting TaskSetFailed Event — taskSetFailed Method

taskSetFailed(
  taskSet: TaskSet,
  reason: String,
  exception: Option[Throwable]): Unit

taskSetFailed simply posts a TaskSetFailed to DAGScheduler’s Internal Event Bus.

Note
The input arguments of taskSetFailed are exactly the arguments of TaskSetFailed.
Note
taskSetFailed is executed when a TaskSetManager is aborted.

Posting ExecutorLost Event — executorLost Method

executorLost(execId: String, reason: ExecutorLossReason): Unit

executorLost simply posts a ExecutorLost event to DAGScheduler’s Internal Event Bus.

Posting ExecutorAdded Event — executorAdded Method

executorAdded(execId: String, host: String): Unit

executorAdded simply posts a ExecutorAdded event to DAGScheduler’s Internal Event Bus.

Posting JobCancelled Event — cancelJob Method

cancelJob(jobId: Int): Unit

cancelJob prints the following INFO message and posts a JobCancelled to DAGScheduler’s Internal Event Bus.

INFO DAGScheduler: Asked to cancel job [id]
Note
cancelJob is called when SparkContext and JobWaiter are requested to cancel a Spark job.

Finding Or Creating Missing Direct Parent ShuffleMapStages (For ShuffleDependencies of Input RDD) — getOrCreateParentStages Internal Method

getOrCreateParentStages(rdd: RDD[_], firstJobId: Int): List[Stage]

getOrCreateParentStages finds all direct parent ShuffleDependencies of the input rdd and then finds ShuffleMapStage stages for each ShuffleDependency.

Note
getOrCreateParentStages is used when DAGScheduler createShuffleMapStage and createResultStage.

Marking Stage Finished — markStageAsFinished Internal Method

markStageAsFinished(stage: Stage, errorMessage: Option[String] = None): Unit
Caution
FIXME

Running Job — runJob Method

runJob[T, U](
  rdd: RDD[T],
  func: (TaskContext, Iterator[T]) => U,
  partitions: Seq[Int],
  callSite: CallSite,
  resultHandler: (Int, U) => Unit,
  properties: Properties): Unit

runJob submits an action job to the DAGScheduler and waits for a result.

Internally, runJob executes submitJob and then waits until a result comes using JobWaiter.

When the job succeeds, you should see the following INFO message in the logs:

INFO Job [jobId] finished: [callSite], took [time] s

When the job fails, you should see the following INFO message in the logs and the exception (that led to the failure) is thrown.

INFO Job [jobId] failed: [callSite], took [time] s
Note
runJob is used when SparkContext runs a job.

Finding or Creating New ShuffleMapStages for ShuffleDependency — getOrCreateShuffleMapStage Internal Method

getOrCreateShuffleMapStage(
  shuffleDep: ShuffleDependency[_, _, _],
  firstJobId: Int): ShuffleMapStage

getOrCreateShuffleMapStage finds or creates the ShuffleMapStage for the input ShuffleDependency.

Internally, getOrCreateShuffleMapStage finds the ShuffleDependency in shuffleIdToMapStage internal registry and returns one when found.

If no ShuffleDependency was available, getOrCreateShuffleMapStage finds all the missing shuffle dependencies and creates corresponding ShuffleMapStage stages (including one for the input shuffleDep).

Note
All the new ShuffleMapStage stages are associated with the input firstJobId.

Creating ShuffleMapStage for ShuffleDependency (Copying Shuffle Map Output Locations From Previous Jobs) — createShuffleMapStage Method

createShuffleMapStage(
  shuffleDep: ShuffleDependency[_, _, _],
  jobId: Int): ShuffleMapStage

createShuffleMapStage creates a ShuffleMapStage for the input ShuffleDependency and jobId (of a ActiveJob) possibly copying shuffle map output locations from previous jobs to avoid recomputing records.

Note
When a ShuffleMapStage is created, the id is generated (using nextStageId internal counter), rdd is from ShuffleDependency, numTasks is the number of partitions in the RDD, all parents are looked up (and possibly created), the jobId is given, callSite is the creationSite of the RDD, and shuffleDep is the input ShuffleDependency.
Note
ShuffleDependency is associated with exactly one RDD[Product2[K, V]].

createShuffleMapStage creates a ShuffleMapStage (with the stage id from nextStageId internal counter).

Note
The RDD of the new ShuffleMapStage is from the input ShuffleDependency.

createShuffleMapStage registers the ShuffleMapStage in stageIdToStage and shuffleIdToMapStage internal registries.

createShuffleMapStage calls updateJobIdStageIdMaps.

DAGScheduler MapOutputTrackerMaster containsShuffle.png
Figure 5. DAGScheduler Asks MapOutputTrackerMaster Whether Shuffle Map Output Is Already Tracked

If however MapOutputTrackerMaster does not track the input ShuffleDependency, you should see the following INFO message in the logs and createShuffleMapStage registers the ShuffleDependency with MapOutputTrackerMaster.

INFO Registering RDD [id] ([creationSite])

createShuffleMapStage returns the new ShuffleMapStage.

Note
createShuffleMapStage is executed only when DAGScheduler finds or creates parent ShuffleMapStage stages for a ShuffleDependency.

Clearing Cache of RDD Block Locations — clearCacheLocs Internal Method

clearCacheLocs(): Unit
Note
DAGScheduler clears the cache while resubmitting failed stages, and as a result of JobSubmitted, MapStageSubmitted, CompletionEvent, ExecutorLost events.

Finding Missing ShuffleDependencies For RDD — getMissingAncestorShuffleDependencies Internal Method

getMissingAncestorShuffleDependencies(rdd: RDD[_]): Stack[ShuffleDependency[_, _, _]]

getMissingAncestorShuffleDependencies finds all missing shuffle dependencies for the given RDD traversing its dependency chain (aka RDD lineage).

Note
A missing shuffle dependency of a RDD is a dependency not registered in shuffleIdToMapStage internal registry.

Internally, getMissingAncestorShuffleDependencies finds direct parent shuffle dependencies of the input RDD and collects the ones that are not registered in shuffleIdToMapStage internal registry. It repeats the process for the RDDs of the parent shuffle dependencies.

Note
getMissingAncestorShuffleDependencies is used when DAGScheduler finds all ShuffleMapStage stages for a ShuffleDependency.

Finding Direct Parent Shuffle Dependencies of RDD — getShuffleDependencies Internal Method

getShuffleDependencies(rdd: RDD[_]): HashSet[ShuffleDependency[_, _, _]]

getShuffleDependencies finds direct parent shuffle dependencies for the given RDD.

spark DAGScheduler getShuffleDependencies.png
Figure 6. getShuffleDependencies Finds Direct Parent ShuffleDependencies (shuffle1 and shuffle2)

Internally, getShuffleDependencies takes the direct shuffle dependencies of the input RDD and direct shuffle dependencies of all the parent non-ShuffleDependencies in the dependency chain (aka RDD lineage).

Note
getShuffleDependencies is used when DAGScheduler finds or creates missing direct parent ShuffleMapStages (for ShuffleDependencies of given RDD) and finds all missing shuffle dependencies for a given RDD.

Failing Job and Single-Job Stages — failJobAndIndependentStages Internal Method

failJobAndIndependentStages(
  job: ActiveJob,
  failureReason: String,
  exception: Option[Throwable] = None): Unit

The internal failJobAndIndependentStages method fails the input job and all the stages that are only used by the job.

Internally, failJobAndIndependentStages uses jobIdToStageIds internal registry to look up the stages registered for the job.

If no stages could be found, you should see the following ERROR message in the logs:

ERROR No stages registered for job [id]

Otherwise, for every stage, failJobAndIndependentStages finds the job ids the stage belongs to.

If no stages could be found or the job is not referenced by the stages, you should see the following ERROR message in the logs:

ERROR Job [id] not registered for stage [id] even though that stage was registered for the job

Only when there is exactly one job registered for the stage and the stage is in RUNNING state (in runningStages internal registry), TaskScheduler is requested to cancel the stage’s tasks and marks the stage finished.

Note
failJobAndIndependentStages is called from handleJobCancellation and abortStage.
Note
failJobAndIndependentStages uses jobIdToStageIds, stageIdToStage, and runningStages internal registries.

Aborting Stage — abortStage Internal Method

abortStage(
  failedStage: Stage,
  reason: String,
  exception: Option[Throwable]): Unit

abortStage is an internal method that finds all the active jobs that depend on the failedStage stage and fails them.

Internally, abortStage looks the failedStage stage up in the internal stageIdToStage registry and exits if there the stage was not registered earlier.

If it was, abortStage finds all the active jobs (in the internal activeJobs registry) with the final stage depending on the failedStage stage.

At this time, the completionTime property (of the failed stage’s StageInfo) is assigned to the current time (millis).

All the active jobs that depend on the failed stage (as calculated above) and the stages that do not belong to other jobs (aka independent stages) are failed (with the failure reason being "Job aborted due to stage failure: [reason]" and the input exception).

If there are no jobs depending on the failed stage, you should see the following INFO message in the logs:

INFO Ignoring failure of [failedStage] because all jobs depending on it are done

Checking Out Stage Dependency on Given Stage — stageDependsOn Method

stageDependsOn(stage: Stage, target: Stage): Boolean

stageDependsOn compares two stages and returns whether the stage depends on target stage (i.e. true) or not (i.e. false).

Note
A stage A depends on stage B if B is among the ancestors of A.

Internally, stageDependsOn walks through the graph of RDDs of the input stage. For every RDD in the RDD’s dependencies (using RDD.dependencies) stageDependsOn adds the RDD of a NarrowDependency to a stack of RDDs to visit while for a ShuffleDependency it finds ShuffleMapStage stages for a ShuffleDependency for the dependency and the stage's first job id that it later adds to a stack of RDDs to visit if the map stage is ready, i.e. all the partitions have shuffle outputs.

After all the RDDs of the input stage are visited, stageDependsOn checks if the target's RDD is among the RDDs of the stage, i.e. whether the stage depends on target stage.

dag-scheduler-event-loop — DAGScheduler’s Internal Event Bus

eventProcessLoop is DAGScheduler’s event bus to which Spark (by submitJob) posts jobs to schedule their execution. Later on, TaskSetManager talks back to DAGScheduler to inform about the status of the tasks using the same "communication channel".

It allows Spark to release the current thread when posting happens and let the event loop handle events on a separate thread - asynchronously.

…​IMAGE…​FIXME

Caution
FIXME statistics? MapOutputStatistics?

Submitting Waiting Child Stages for Execution — submitWaitingChildStages Internal Method

submitWaitingChildStages(parent: Stage): Unit

submitWaitingChildStages submits for execution all waiting stages for which the input parent Stage is the direct parent.

Note
Waiting stages are the stages registered in waitingStages internal registry.

When executed, you should see the following TRACE messages in the logs:

TRACE DAGScheduler: Checking if any dependencies of [parent] are now runnable
TRACE DAGScheduler: running: [runningStages]
TRACE DAGScheduler: waiting: [waitingStages]
TRACE DAGScheduler: failed: [failedStages]

submitWaitingChildStages finds child stages of the input parent stage, removes them from waitingStages internal registry, and submits one by one sorted by their job ids.

Note
submitWaitingChildStages is executed when DAGScheduler submits missing tasks for stage and handles successful ShuffleMapTask completion.

Submitting Stage or Its Missing Parents for Execution — submitStage Internal Method

submitStage(stage: Stage)

submitStage is an internal method that DAGScheduler uses to submit the input stage or its missing parents (if there any stages not computed yet before the input stage could).

Note
submitStage is also used to resubmit failed stages.

submitStage recursively submits any missing parents of the stage.

Internally, submitStage first finds the earliest-created job id that needs the stage.

Note
A stage itself tracks the jobs (their ids) it belongs to (using the internal jobIds registry).

The following steps depend on whether there is a job or not.

If there are no jobs that require the stage, submitStage aborts it with the reason:

No active job for stage [id]

If however there is a job for the stage, you should see the following DEBUG message in the logs:

DEBUG DAGScheduler: submitStage([stage])

submitStage checks the status of the stage and continues when it was not recorded in waiting, running or failed internal registries. It simply exits otherwise.

With the stage ready for submission, submitStage calculates the list of missing parent stages of the stage (sorted by their job ids). You should see the following DEBUG message in the logs:

DEBUG DAGScheduler: missing: [missing]

When the stage has no parent stages missing, you should see the following INFO message in the logs:

INFO DAGScheduler: Submitting [stage] ([stage.rdd]), which has no missing parents

submitStage submits the stage (with the earliest-created job id) and finishes.

If however there are missing parent stages for the stage, submitStage submits all the parent stages, and the stage is recorded in the internal waitingStages registry.

Note
submitStage is executed when DAGScheduler submits missing parent map stages (of a stage) recursively or waiting child stages, resubmits failed stages, and handles JobSubmitted, MapStageSubmitted, or CompletionEvent events.

Fault recovery - stage attempts

A single stage can be re-executed in multiple attempts due to fault recovery. The number of attempts is configured (FIXME).

If TaskScheduler reports that a task failed because a map output file from a previous stage was lost, the DAGScheduler resubmits the lost stage. This is detected through a CompletionEvent with FetchFailed, or an ExecutorLost event. DAGScheduler will wait a small amount of time to see whether other nodes or tasks fail, then resubmit TaskSets for any lost stage(s) that compute the missing tasks.

Please note that tasks from the old attempts of a stage could still be running.

A stage object tracks multiple StageInfo objects to pass to Spark listeners or the web UI.

The latest StageInfo for the most recent attempt for a stage is accessible through latestInfo.

Preferred Locations

DAGScheduler computes where to run each task in a stage based on the preferred locations of its underlying RDDs, or the location of cached or shuffle data.

Adaptive Query Planning / Adaptive Scheduling

See SPARK-9850 Adaptive execution in Spark for the design document. The work is currently in progress.

DAGScheduler.submitMapStage method is used for adaptive query planning, to run map stages and look at statistics about their outputs before submitting downstream stages.

ScheduledExecutorService daemon services

DAGScheduler uses the following ScheduledThreadPoolExecutors (with the policy of removing cancelled tasks from a work queue at time of cancellation):

They are created using ThreadUtils.newDaemonSingleThreadScheduledExecutor method that uses Guava DSL to instantiate a ThreadFactory.

Finding Missing Parent ShuffleMapStages For Stage — getMissingParentStages Internal Method

getMissingParentStages(stage: Stage): List[Stage]

getMissingParentStages finds missing parent ShuffleMapStages in the dependency graph of the input stage (using the breadth-first search algorithm).

Internally, getMissingParentStages starts with the stage's RDD and walks up the tree of all parent RDDs to find uncached partitions.

Note
A Stage tracks the associated RDD using rdd property.
Note
An uncached partition of a RDD is a partition that has Nil in the internal registry of partition locations per RDD (which results in no RDD blocks in any of the active BlockManagers on executors).

getMissingParentStages traverses the parent dependencies of the RDD and acts according to their type, i.e. ShuffleDependency or NarrowDependency.

Note
ShuffleDependency and NarrowDependency are the main top-level Dependencies.

For each NarrowDependency, getMissingParentStages simply marks the corresponding RDD to visit and moves on to a next dependency of a RDD or works on another unvisited parent RDD.

Note
NarrowDependency is a RDD dependency that allows for pipelined execution.

getMissingParentStages focuses on ShuffleDependency dependencies.

Note
ShuffleDependency is a RDD dependency that represents a dependency on the output of a ShuffleMapStage, i.e. shuffle map stage.

For each ShuffleDependency, getMissingParentStages finds ShuffleMapStage stages. If the ShuffleMapStage is not available, it is added to the set of missing (map) stages.

Note
A ShuffleMapStage is available when all its partitions are computed, i.e. results are available (as blocks).
Caution
FIXME…​IMAGE with ShuffleDependencies queried
Note
getMissingParentStages is used when DAGScheduler submits missing parent ShuffleMapStages (of a stage) and handles JobSubmitted and MapStageSubmitted events.

Submitting Missing Tasks for Stage (and Job) — submitMissingTasks Internal Method

submitMissingTasks(stage: Stage, jobId: Int): Unit

submitMissingTasks is an internal method that…​FIXME

When submitMissingTasks is executed, you should see the following DEBUG message in the logs:

DEBUG DAGScheduler: submitMissingTasks([stage])

The input stage's pendingPartitions internal field is cleared (it is later filled out with the partitions to run tasks for).

submitMissingTasks requests the stage for missing partitions, i.e. the indices of the partitions to compute.

submitMissingTasks marks the stage as running (i.e. adds it to runningStages internal registry).

Note
The input maxPartitionId argument handed over to OutputCommitCoordinator depends on the type of the stage, i.e. ShuffleMapStage or ResultStage. ShuffleMapStage tracks the number of partitions itself (as numPartitions property) while ResultStage uses the internal RDD to find out the number.

For the missing partitions, submitMissingTasks computes their task locality preferences, i.e. pairs of missing partition ids and their task locality information. HERE NOTE: The locality information of a RDD is called preferred locations.

In case of non-fatal exceptions at this time (while getting the locality information), submitMissingTasks creates a new stage attempt.

Note
A stage attempt is an internal property of a stage.

Despite the failure to submit any tasks, submitMissingTasks does announce that at least there was an attempt on LiveListenerBus by posting a SparkListenerStageSubmitted message.

Note
The Spark application’s LiveListenerBus is given when DAGScheduler is created.

submitMissingTasks then aborts the stage (with the reason being "Task creation failed" followed by the exception).

The stage is removed from the internal runningStages collection of stages and submitMissingTasks exits.

When no exception was thrown (while computing the locality information for tasks), submitMissingTasks creates a new stage attempt and announces it on LiveListenerBus by posting a SparkListenerStageSubmitted message.

Note
Yes, that is correct. Whether there was a task submission failure or not, submitMissingTasks creates a new stage attempt and posts a SparkListenerStageSubmitted. That makes sense, doesn’t it?

At that time, submitMissingTasks serializes the RDD (of the stage for which tasks are submitted for) and, depending on the type of the stage, the ShuffleDependency (for ShuffleMapStage) or the function (for ResultStage).

Note
submitMissingTasks uses a closure Serializer that DAGScheduler creates for the entire lifetime when it is created. The closure serializer is available through SparkEnv.

The serialized so-called task binary bytes are "wrapped" as a broadcast variable (to make it available for executors to execute later on).

Note
That exact moment should make clear how important broadcast variables are for Spark itself that you, a Spark developer, can use, too, to distribute data across the nodes in a Spark application in a very efficient way.

Any NotSerializableException exceptions lead to aborting the stage (with the reason being "Task not serializable: [exception]") and removing the stage from the internal runningStages collection of stages. submitMissingTasks exits.

Any non-fatal exceptions lead to aborting the stage (with the reason being "Task serialization failed" followed by the exception) and removing the stage from the internal runningStages collection of stages. submitMissingTasks exits.

With no exceptions along the way, submitMissingTasks computes a collection of tasks to execute for the missing partitions (of the stage).

submitMissingTasks creates a ShuffleMapTask or ResultTask for every missing partition of the stage being ShuffleMapStage or ResultStage, respectively. submitMissingTasks uses the preferred locations (computed earlier) per partition.

Caution
FIXME Image with creating tasks for partitions in the stage.

Any non-fatal exceptions lead to aborting the stage (with the reason being "Task creation failed" followed by the exception) and removing the stage from the internal runningStages collection of stages. submitMissingTasks exits.

If there are tasks to submit for execution (i.e. there are missing partitions in the stage), you should see the following INFO message in the logs:

INFO DAGScheduler: Submitting [size] missing tasks from [stage] ([rdd])

submitMissingTasks records the partitions (of the tasks) in the stage's pendingPartitions property.

Note
pendingPartitions property of the stage was cleared when submitMissingTasks started.

You should see the following DEBUG message in the logs:

DEBUG DAGScheduler: New pending partitions: [pendingPartitions]

submitMissingTasks submits the tasks to TaskScheduler for execution (with the id of the stage, attempt id, the input jobId, and the properties of the ActiveJob with jobId).

Note
A TaskScheduler was given when DAGScheduler was created.
Caution
FIXME What are the ActiveJob properties for? Where are they used?

submitMissingTasks records the submission time in the stage’s StageInfo and exits.

If however there are no tasks to submit for execution, submitMissingTasks marks the stage as finished (with no errorMessage).

You should see a DEBUG message that varies per the type of the input stage which are:

DEBUG DAGScheduler: Stage [stage] is actually done; (available: [isAvailable],available outputs: [numAvailableOutputs],partitions: [numPartitions])

or

DEBUG DAGScheduler: Stage [stage] is actually done; (partitions: [numPartitions])

for ShuffleMapStage and ResultStage, respectively.

In the end, with no tasks to submit for execution, submitMissingTasks submits waiting child stages for execution and exits.

Note
submitMissingTasks is called when DAGScheduler submits a stage for execution.

Computing Preferred Locations for Missing Partitions — getPreferredLocs Method

getPreferredLocs(rdd: RDD[_], partition: Int): Seq[TaskLocation]

getPreferredLocs is simply an alias for the internal (recursive) getPreferredLocsInternal.

Finding BlockManagers (Executors) for Cached RDD Partitions (aka Block Location Discovery) — getCacheLocs Internal Method

getCacheLocs(rdd: RDD[_]): IndexedSeq[Seq[TaskLocation]]

getCacheLocs gives TaskLocations (block locations) for the partitions of the input rdd. getCacheLocs caches lookup results in cacheLocs internal registry.

Note
The size of the collection from getCacheLocs is exactly the number of partitions in rdd RDD.
Note
The size of every TaskLocation collection (i.e. every entry in the result of getCacheLocs) is exactly the number of blocks managed using BlockManagers on executors.

Internally, getCacheLocs finds rdd in the cacheLocs internal registry (of partition locations per RDD).

If rdd is not in cacheLocs internal registry, getCacheLocs branches per its storage level.

For NONE storage level (i.e. no caching), the result is an empty locations (i.e. no location preference).

For other non-NONE storage levels, getCacheLocs requests BlockManagerMaster for block locations that are then mapped to TaskLocations with the hostname of the owning BlockManager for a block (of a partition) and the executor id.

Note
getCacheLocs uses BlockManagerMaster that was defined when DAGScheduler was created.

getCacheLocs records the computed block locations per partition (as TaskLocation) in cacheLocs internal registry.

Note
getCacheLocs requests locations from BlockManagerMaster using RDDBlockId with the RDD id and the partition indices (which implies that the order of the partitions matters to request proper blocks).
Note
DAGScheduler uses TaskLocations (with host and executor) while BlockManagerMaster uses BlockManagerId (to track similar information, i.e. block locations).
Note
getCacheLocs is used when DAGScheduler finds missing parent MapStages and getPreferredLocsInternal.

Finding Placement Preferences for RDD Partition (recursively) — getPreferredLocsInternal Internal Method

getPreferredLocsInternal(
  rdd: RDD[_],
  partition: Int,
  visited: HashSet[(RDD[_], Int)]): Seq[TaskLocation]

getPreferredLocsInternal first finds the TaskLocations for the partition of the rdd (using cacheLocs internal cache) and returns them.

Otherwise, if not found, getPreferredLocsInternal requests rdd for the preferred locations of partition and returns them.

Note
Preferred locations of the partitions of a RDD are also called placement preferences or locality preferences.

Otherwise, if not found, getPreferredLocsInternal finds the first parent NarrowDependency and (recursively) finds TaskLocations.

If all the attempts fail to yield any non-empty result, getPreferredLocsInternal returns an empty collection of TaskLocations.

Note
getPreferredLocsInternal is used exclusively when DAGScheduler computes preferred locations for missing partitions.

Stopping DAGScheduler — stop Method

stop(): Unit

stop stops the internal dag-scheduler-message thread pool, dag-scheduler-event-loop, and TaskScheduler.

DAGSchedulerSource Metrics Source

DAGScheduler uses Spark Metrics System (via DAGSchedulerSource) to report metrics about internal status.

Caution
FIXME What is DAGSchedulerSource?

The name of the source is DAGScheduler.

It emits the following numbers:

  • stage.failedStages - the number of failed stages

  • stage.runningStages - the number of running stages

  • stage.waitingStages - the number of waiting stages

  • job.allJobs - the number of all jobs

  • job.activeJobs - the number of active jobs

Updating Accumulators with Partial Values from Completed Tasks — updateAccumulators Internal Method

updateAccumulators(event: CompletionEvent): Unit

The private updateAccumulators method merges the partial values of accumulators from a completed task into their "source" accumulators on the driver.

Note
It is called by handleTaskCompletion.

For each AccumulableInfo in the CompletionEvent, a partial value from a task is obtained (from AccumulableInfo.update) and added to the driver’s accumulator (using Accumulable.++= method).

For named accumulators with the update value being a non-zero value, i.e. not Accumulable.zero:

  • stage.latestInfo.accumulables for the AccumulableInfo.id is set

  • CompletionEvent.taskInfo.accumulables has a new AccumulableInfo added.

Caution
FIXME Where are Stage.latestInfo.accumulables and CompletionEvent.taskInfo.accumulables used?

Settings

Table 2. Spark Properties
Spark Property Default Value Description

spark.test.noStageRetry

false

When enabled (i.e. true), task failures with FetchFailed exceptions will not cause stage retries, in order to surface the problem. Used for testing.

results matching ""

    No results matching ""