[Refactor] Extend REPL to support external metadata storage and data …

…storage (#381) * REPL refactor Signed-off-by: Louis Chu <[email protected]> * Ignore UT Signed-off-by: Louis Chu <[email protected]> * Add seqNo and primaryTerm to session object Signed-off-by: Louis Chu <[email protected]> * Address os client concurrency limitation Signed-off-by: Louis Chu <[email protected]> * Add interface for query execution Signed-off-by: Louis Chu <[email protected]> * Clean up logs and add UTs Signed-off-by: Louis Chu <[email protected]> * Clean up and fix UTs Signed-off-by: Louis Chu <[email protected]> --------- Signed-off-by: Louis Chu <[email protected]>
opensearch-project · Aug 28, 2024 · aa509ac · aa509ac
1 parent 7fd9223
commit aa509ac
Show file tree

Hide file tree

Showing 21 changed files with 1,328 additions and 748 deletions.
diff --git a/flint-commons/src/main/scala/org/apache/spark/sql/QueryResultWriter.scala b/flint-commons/src/main/scala/org/apache/spark/sql/QueryResultWriter.scala
@@ -0,0 +1,20 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+package org.apache.spark.sql
+
+import org.opensearch.flint.common.model.FlintStatement
+
+/**
+ * Trait for writing the result of a query execution to an external data storage.
+ */
+trait QueryResultWriter {
+
+  /**
+   * Writes the given DataFrame, which represents the result of a query execution, to an external
+   * data storage based on the provided FlintStatement metadata.
+   */
+  def writeDataFrame(dataFrame: DataFrame, flintStatement: FlintStatement): Unit
+}
diff --git a/flint-commons/src/main/scala/org/apache/spark/sql/SessionManager.scala b/flint-commons/src/main/scala/org/apache/spark/sql/SessionManager.scala
@@ -0,0 +1,43 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+package org.apache.spark.sql
+
+import org.opensearch.flint.common.model.{FlintStatement, InteractiveSession}
+
+import org.apache.spark.sql.SessionUpdateMode.SessionUpdateMode
+
+/**
+ * Trait defining the interface for managing interactive sessions.
+ */
+trait SessionManager {
+
+  /**
+   * Retrieves metadata about the session manager.
+   */
+  def getSessionContext: Map[String, Any]
+
+  /**
+   * Fetches the details of a specific session.
+   */
+  def getSessionDetails(sessionId: String): Option[InteractiveSession]
+
+  /**
+   * Updates the details of a specific session.
+   */
+  def updateSessionDetails(
+      sessionDetails: InteractiveSession,
+      updateMode: SessionUpdateMode): Unit
+
+  /**
+   * Records a heartbeat for a specific session to indicate it is still active.
+   */
+  def recordHeartbeat(sessionId: String): Unit
+}
+
+object SessionUpdateMode extends Enumeration {
+  type SessionUpdateMode = Value
+  val UPDATE, UPSERT, UPDATE_IF = Value
+}
diff --git a/flint-commons/src/main/scala/org/apache/spark/sql/StatementExecutionManager.scala b/flint-commons/src/main/scala/org/apache/spark/sql/StatementExecutionManager.scala
@@ -0,0 +1,42 @@
+/*
+ * Copyright OpenSearch Contributors
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+package org.apache.spark.sql
+
+import org.opensearch.flint.common.model.FlintStatement
+
+/**
+ * Trait defining the interface for managing FlintStatement execution. For example, in FlintREPL,
+ * multiple FlintStatements are running in a micro-batch within same session.
+ *
+ * This interface can also apply to other spark entry point like FlintJob.
+ */
+trait StatementExecutionManager {
+
+  /**
+   * Prepares execution of each individual statement
+   */
+  def prepareStatementExecution(): Either[String, Unit]
+
+  /**
+   * Executes a specific statement and returns the spark dataframe
+   */
+  def executeStatement(statement: FlintStatement): DataFrame
+
+  /**
+   * Retrieves the next statement to be executed.
+   */
+  def getNextStatement(): Option[FlintStatement]
+
+  /**
+   * Updates a specific statement.
+   */
+  def updateStatement(statement: FlintStatement): Unit
+
+  /**
+   * Terminates the statement lifecycle.
+   */
+  def terminateStatementsExecution(): Unit
+}
diff --git a/flint-commons/src/main/scala/org/opensearch/flint/common/model/FlintStatement.scala b/flint-commons/src/main/scala/org/opensearch/flint/common/model/FlintStatement.scala
@@ -65,7 +65,7 @@ class FlintStatement(
 
   // Does not include context, which could contain sensitive information.
   override def toString: String =
-    s"FlintStatement(state=$state, query=$query, statementId=$statementId, queryId=$queryId, submitTime=$submitTime, error=$error)"
+    s"FlintStatement(state=$state, statementId=$statementId, queryId=$queryId, submitTime=$submitTime, error=$error)"
 }
 
 object FlintStatement {

diff --git a/flint-commons/src/main/scala/org/opensearch/flint/common/model/InteractiveSession.scala b/flint-commons/src/main/scala/org/opensearch/flint/common/model/InteractiveSession.scala
@@ -14,6 +14,8 @@ import org.json4s.JsonAST.{JArray, JString}
 import org.json4s.native.JsonMethods.parse
 import org.json4s.native.Serialization
 
+import org.apache.spark.internal.Logging
+
 object SessionStates {
   val RUNNING = "running"
   val DEAD = "dead"
@@ -52,7 +54,8 @@ class InteractiveSession(
     val excludedJobIds: Seq[String] = Seq.empty[String],
     val error: Option[String] = None,
     sessionContext: Map[String, Any] = Map.empty[String, Any])
-    extends ContextualDataStore {
+    extends ContextualDataStore
+    with Logging {
   context = sessionContext // Initialize the context from the constructor
 
   def running(): Unit = state = SessionStates.RUNNING
@@ -96,6 +99,7 @@ object InteractiveSession {
     // Replace extractOpt with jsonOption and map
     val excludeJobIds: Seq[String] = meta \ "excludeJobIds" match {
       case JArray(lst) => lst.map(_.extract[String])
+      case JString(s) => Seq(s)
       case _ => Seq.empty[String]
     }
 

diff --git a/flint-core/src/main/scala/org/opensearch/flint/core/storage/OpenSearchReader.java b/flint-core/src/main/scala/org/opensearch/flint/core/storage/OpenSearchReader.java
@@ -22,7 +22,6 @@
  * Abstract OpenSearch Reader.
  */
 public abstract class OpenSearchReader implements FlintReader {
-
   @VisibleForTesting
   /** Search request source builder. */
   public final SearchRequest searchRequest;

diff --git a/...t-spark-integration/src/main/scala/org/apache/spark/sql/flint/config/FlintSparkConf.scala b/...t-spark-integration/src/main/scala/org/apache/spark/sql/flint/config/FlintSparkConf.scala
@@ -235,6 +235,15 @@ object FlintSparkConf {
     FlintConfig("spark.metadata.accessAWSCredentialsProvider")
       .doc("AWS credentials provider for metadata access permission")
       .createOptional()
+  val CUSTOM_SESSION_MANAGER =
+    FlintConfig("spark.flint.job.customSessionManager")
+      .createOptional()
+  val CUSTOM_STATEMENT_MANAGER =
+    FlintConfig("spark.flint.job.customStatementManager")
+      .createOptional()
+  val CUSTOM_QUERY_RESULT_WRITER =
+    FlintConfig("spark.flint.job.customQueryResultWriter")
+      .createOptional()
 }
 
 /**

diff --git a/integ-test/src/integration/scala/org/apache/spark/sql/FlintJobITSuite.scala b/integ-test/src/integration/scala/org/apache/spark/sql/FlintJobITSuite.scala
@@ -98,9 +98,15 @@ class FlintJobITSuite extends FlintSparkSuite with JobTest {
        * JobOperator instance to accommodate specific runtime requirements.
        */
       val job =
-        JobOperator(spark, query, dataSourceName, resultIndex, true, streamingRunningCount)
-      job.envinromentProvider = new MockEnvironment(
-        Map("SERVERLESS_EMR_JOB_ID" -> jobRunId, "SERVERLESS_EMR_VIRTUAL_CLUSTER_ID" -> appId))
+        JobOperator(
+          appId,
+          jobRunId,
+          spark,
+          query,
+          dataSourceName,
+          resultIndex,
+          true,
+          streamingRunningCount)
       job.terminateJVM = false
       job.start()
     }

diff --git a/integ-test/src/integration/scala/org/apache/spark/sql/FlintREPLITSuite.scala b/integ-test/src/integration/scala/org/apache/spark/sql/FlintREPLITSuite.scala
@@ -169,7 +169,7 @@ class FlintREPLITSuite extends SparkFunSuite with OpenSearchSuite with JobTest {
         "spark.flint.job.queryLoopExecutionFrequency",
         queryLoopExecutionFrequency.toString)
 
-      FlintREPL.envinromentProvider = new MockEnvironment(
+      FlintREPL.environmentProvider = new MockEnvironment(
         Map("SERVERLESS_EMR_JOB_ID" -> jobRunId, "SERVERLESS_EMR_VIRTUAL_CLUSTER_ID" -> appId))
       FlintREPL.enableHiveSupport = false
       FlintREPL.terminateJVM = false

diff --git a/integ-test/src/integration/scala/org/opensearch/flint/OpenSearchSuite.scala b/integ-test/src/integration/scala/org/opensearch/flint/OpenSearchSuite.scala
@@ -145,7 +145,6 @@ trait OpenSearchSuite extends BeforeAndAfterAll {
 
       val response =
         openSearchClient.bulk(request, RequestOptions.DEFAULT)
-
       assume(
         !response.hasFailures,
         s"bulk index docs to $index failed: ${response.buildFailureMessage()}")

diff --git a/spark-sql-application/src/main/scala/org/apache/spark/sql/CommandContext.scala b/spark-sql-application/src/main/scala/org/apache/spark/sql/CommandContext.scala
@@ -11,14 +11,13 @@ import scala.concurrent.duration.Duration
 import org.opensearch.flint.core.storage.{FlintReader, OpenSearchUpdater}
 
 case class CommandContext(
+    applicationId: String,
+    jobId: String,
     spark: SparkSession,
     dataSource: String,
-    resultIndex: String,
     sessionId: String,
-    flintSessionIndexUpdater: OpenSearchUpdater,
-    osClient: OSClient,
-    sessionIndex: String,
-    jobId: String,
+    sessionManager: SessionManager,
+    queryResultWriter: QueryResultWriter,
     queryExecutionTimeout: Duration,
     inactivityLimitMillis: Long,
     queryWaitTimeMillis: Long,

diff --git a/spark-sql-application/src/main/scala/org/apache/spark/sql/CommandState.scala b/spark-sql-application/src/main/scala/org/apache/spark/sql/CommandState.scala
@@ -12,7 +12,6 @@ import org.opensearch.flint.core.storage.FlintReader
 case class CommandState(
     recordedLastActivityTime: Long,
     recordedVerificationResult: VerificationResult,
-    flintReader: FlintReader,
-    futureMappingCheck: Future[Either[String, Unit]],
+    futurePrepareQueryExecution: Future[Either[String, Unit]],
     executionContext: ExecutionContextExecutor,
     recordedLastCanPickCheckTime: Long)
diff --git a/spark-sql-application/src/main/scala/org/apache/spark/sql/FlintJob.scala b/spark-sql-application/src/main/scala/org/apache/spark/sql/FlintJob.scala
@@ -29,7 +29,7 @@ import org.apache.spark.sql.types._
  */
 object FlintJob extends Logging with FlintJobExecutor {
   def main(args: Array[String]): Unit = {
-    val (queryOption, resultIndex) = parseArgs(args)
+    val (queryOption, resultIndexOption) = parseArgs(args)
 
     val conf = createSparkConf()
     val jobType = conf.get("spark.flint.job.type", "batch")
@@ -41,6 +41,9 @@ object FlintJob extends Logging with FlintJobExecutor {
     if (query.isEmpty) {
       logAndThrow(s"Query undefined for the ${jobType} job.")
     }
+    if (resultIndexOption.isEmpty) {
+      logAndThrow("resultIndex is not set")
+    }
     // https://github.com/opensearch-project/opensearch-spark/issues/138
     /*
      * To execute queries such as `CREATE SKIPPING INDEX ON my_glue1.default.http_logs_plain (`@timestamp` VALUE_SET) WITH (auto_refresh = true)`,
@@ -52,13 +55,19 @@ object FlintJob extends Logging with FlintJobExecutor {
     conf.set("spark.sql.defaultCatalog", dataSource)
     configDYNMaxExecutors(conf, jobType)
 
+    val applicationId =
+      environmentProvider.getEnvVar("SERVERLESS_EMR_VIRTUAL_CLUSTER_ID", "unknown")
+    val jobId = environmentProvider.getEnvVar("SERVERLESS_EMR_JOB_ID", "unknown")
+
     val streamingRunningCount = new AtomicInteger(0)
     val jobOperator =
       JobOperator(
+        applicationId,
+        jobId,
         createSparkSession(conf),
         query,
         dataSource,
-        resultIndex,
+        resultIndexOption.get,
         jobType.equalsIgnoreCase("streaming"),
         streamingRunningCount)
     registerGauge(MetricConstants.STREAMING_RUNNING_METRIC, streamingRunningCount)

diff --git a/spark-sql-application/src/main/scala/org/apache/spark/sql/FlintJobExecutor.scala b/spark-sql-application/src/main/scala/org/apache/spark/sql/FlintJobExecutor.scala
@@ -41,7 +41,7 @@ trait FlintJobExecutor {
 
   var currentTimeProvider: TimeProvider = new RealTimeProvider()
   var threadPoolFactory: ThreadPoolFactory = new DefaultThreadPoolFactory()
-  var envinromentProvider: EnvironmentProvider = new RealEnvironment()
+  var environmentProvider: EnvironmentProvider = new RealEnvironment()
   var enableHiveSupport: Boolean = true
   // termiante JVM in the presence non-deamon thread before exiting
   var terminateJVM = true
@@ -190,6 +190,7 @@ trait FlintJobExecutor {
     }
   }
 
+  // scalastyle:off
   /**
    * Create a new formatted dataframe with json result, json schema and EMR_STEP_ID.
    *
@@ -201,6 +202,8 @@ trait FlintJobExecutor {
    *   dataframe with result, schema and emr step id
    */
   def getFormattedData(
+      applicationId: String,
+      jobId: String,
       result: DataFrame,
       spark: SparkSession,
       dataSource: String,
@@ -231,14 +234,13 @@ trait FlintJobExecutor {
     // after consumed the query result. Streaming query shuffle data is cleaned after each
     // microBatch execution.
     cleaner.cleanUp(spark)
-
     // Create the data rows
     val rows = Seq(
       (
         resultToSave,
         resultSchemaToSave,
-        envinromentProvider.getEnvVar("SERVERLESS_EMR_JOB_ID", "unknown"),
-        envinromentProvider.getEnvVar("SERVERLESS_EMR_VIRTUAL_CLUSTER_ID", "unknown"),
+        jobId,
+        applicationId,
         dataSource,
         "SUCCESS",
         "",
@@ -254,6 +256,8 @@ trait FlintJobExecutor {
   }
 
   def constructErrorDF(
+      applicationId: String,
+      jobId: String,
       spark: SparkSession,
       dataSource: String,
       status: String,
@@ -270,8 +274,8 @@ trait FlintJobExecutor {
       (
         null,
         null,
-        envinromentProvider.getEnvVar("SERVERLESS_EMR_JOB_ID", "unknown"),
-        envinromentProvider.getEnvVar("SERVERLESS_EMR_VIRTUAL_CLUSTER_ID", "unknown"),
+        jobId,
+        applicationId,
         dataSource,
         status.toUpperCase(Locale.ROOT),
         error,
@@ -396,6 +400,8 @@ trait FlintJobExecutor {
   }
 
   def executeQuery(
+      applicationId: String,
+      jobId: String,
       spark: SparkSession,
       query: String,
       dataSource: String,
@@ -409,6 +415,8 @@ trait FlintJobExecutor {
     val result: DataFrame = spark.sql(query)
     // Get Data
     getFormattedData(
+      applicationId,
+      jobId,
       result,
       spark,
       dataSource,
@@ -493,16 +501,21 @@ trait FlintJobExecutor {
     }
   }
 
-  def parseArgs(args: Array[String]): (Option[String], String) = {
+  /**
+   * Before OS 2.13, there are two arguments from entry point: query and result index Starting
+   * from OS 2.13, query is optional for FlintREPL And since Flint 0.5, result index is also
+   * optional for non-OpenSearch result persist
+   */
+  def parseArgs(args: Array[String]): (Option[String], Option[String]) = {
     args match {
+      case Array() =>
+        (None, None)
       case Array(resultIndex) =>
-        (None, resultIndex) // Starting from OS 2.13, resultIndex is the only argument
+        (None, Some(resultIndex))
       case Array(query, resultIndex) =>
-        (
-          Some(query),
-          resultIndex
-        ) // Before OS 2.13, there are two arguments, the second one is resultIndex
-      case _ => logAndThrow("Unsupported number of arguments. Expected 1 or 2 arguments.")
+        (Some(query), Some(resultIndex))
+      case _ =>
+        logAndThrow("Unsupported number of arguments. Expected no more than two arguments.")
     }
   }