Skip to content

OperatorStateMetadata with HDFSMetadataLog #10

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
ericm-db wants to merge 9 commits into from
Closed

OperatorStateMetadata with HDFSMetadataLog #10

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
d9d9d99
adding OperatorStateMetadataLog
ericm-db Jun 6, 2024
838c239
moving it to streamexecution
ericm-db Jun 6, 2024
67fe4bf
adding purging logic
ericm-db Jun 6, 2024
d24677b
it writes operatorStateMetadata out
ericm-db Jun 7, 2024
9f5b8f4
new constructor
ericm-db Jun 7, 2024
be7f2d0
only writing out metadatas for new runId
ericm-db Jun 10, 2024
375b980
using match statement instead
ericm-db Jun 10, 2024
c058c76
adding purge methods
ericm-db Jun 10, 2024
f0ccc64
removing accumulator
ericm-db Jun 13, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
moving it to streamexecution
  • Loading branch information
@ericm-db
ericm-db committed Jun 6, 2024
commit 838c239c02becdaced126014f18abc3dce9a8000
20 changes: 2 additions & 18 deletions sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/MicroBatchExecution.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,7 +31,7 @@ import org.apache.spark.sql.catalyst.util.truncatedString
import org.apache.spark.sql.connector.catalog.{SupportsRead, SupportsWrite, TableCapability}
import org.apache.spark.sql.connector.read.streaming.{MicroBatchStream, Offset => OffsetV2, ReadLimit, SparkDataStream, SupportsAdmissionControl, SupportsTriggerAvailableNow}
import org.apache.spark.sql.errors.QueryExecutionErrors
import org.apache.spark.sql.execution.{SparkPlan, SQLExecution}
import org.apache.spark.sql.execution.SQLExecution
import org.apache.spark.sql.execution.datasources.LogicalRelation
import org.apache.spark.sql.execution.datasources.v2.{DataSourceV2Relation, StreamingDataSourceV2Relation, StreamingDataSourceV2ScanRelation, StreamWriterCommitProgress, WriteToDataSourceV2Exec}
import org.apache.spark.sql.execution.streaming.sources.{WriteToMicroBatchDataSource, WriteToMicroBatchDataSourceV1}
Expand Down Expand Up @@ -88,22 +88,6 @@ class MicroBatchExecution(

@volatile protected[sql] var triggerExecutor: TriggerExecutor = _

private lazy val operatorStateMetadatas: Map[Long, OperatorStateMetadataLog] = {
populateOperatorStateMetadatas(getLatestExecutionContext().executionPlan.executedPlan)
}

private def populateOperatorStateMetadatas(plan: SparkPlan):
Map[Long, OperatorStateMetadataLog] = {
plan.flatMap {
case s: StateStoreWriter => s.stateInfo.map { info =>
val metadataPath = s.metadataFilePath()
info.operatorId -> new OperatorStateMetadataLog(sparkSession,
metadataPath.toString)
}
case _ => Seq.empty
}.toMap
}

protected def getTrigger(): TriggerExecutor = {
assert(sources.nonEmpty, "sources should have been retrieved from the plan!")
trigger match {
Expand Down Expand Up @@ -922,7 +906,7 @@ class MicroBatchExecution(
case s: StateStoreWriter =>
val metadata = s.operatorStateMetadata()
val id = metadata.operatorInfo.operatorId
val metadataFile = operatorStateMetadatas(id)
val metadataFile = operatorStateMetadataLogs(id)
if (!metadataFile.add(execCtx.batchId, metadata)) {
throw QueryExecutionErrors.concurrentStreamLogUpdate(execCtx.batchId)
}
Expand Down
17 changes: 17 additions & 0 deletions sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/StreamExecution.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -40,6 +40,7 @@ import org.apache.spark.sql.catalyst.streaming.InternalOutputModes._
import org.apache.spark.sql.connector.catalog.{SupportsWrite, Table}
import org.apache.spark.sql.connector.read.streaming.{Offset => OffsetV2, ReadLimit, SparkDataStream}
import org.apache.spark.sql.connector.write.{LogicalWriteInfoImpl, SupportsTruncate, Write}
import org.apache.spark.sql.execution.SparkPlan
import org.apache.spark.sql.execution.command.StreamingExplainCommand
import org.apache.spark.sql.execution.streaming.sources.ForeachBatchUserFuncException
import org.apache.spark.sql.internal.SQLConf
Expand Down Expand Up @@ -239,6 +240,22 @@ abstract class StreamExecution(
*/
val commitLog = new CommitLog(sparkSession, checkpointFile("commits"))

lazy val operatorStateMetadataLogs: Map[Long, OperatorStateMetadataLog] = {
populateOperatorStateMetadatas(getLatestExecutionContext().executionPlan.executedPlan)
}

private def populateOperatorStateMetadatas(
plan: SparkPlan): Map[Long, OperatorStateMetadataLog] = {
plan.flatMap {
case s: StateStoreWriter => s.stateInfo.map { info =>
val metadataPath = s.metadataFilePath()
info.operatorId -> new OperatorStateMetadataLog(sparkSession,
metadataPath.toString)
}
case _ => Seq.empty
}.toMap
}

/** Whether all fields of the query have been initialized */
private def isInitialized: Boolean = state.get != INITIALIZING

Expand Down