Add base session created in SparkConnectService

Author: garlandz-db

sql/connect/server/src/main/scala/org/apache/spark/sql/connect/service/SparkConnectService.scala

@@ -436,6 +436,7 @@ object SparkConnectService extends Logging {
       return
     }
 
+    sessionManager.initializeBaseSession(sc)
     startGRPCService()
     createListenerAndUI(sc)
 

sql/connect/server/src/main/scala/org/apache/spark/sql/connect/service/SparkConnectSessionManager.scala

@@ -27,7 +27,7 @@ import scala.util.control.NonFatal
 
 import com.google.common.cache.CacheBuilder
 
-import org.apache.spark.{SparkEnv, SparkSQLException}
+import org.apache.spark.{SparkContext, SparkEnv, SparkSQLException}
 import org.apache.spark.internal.Logging
 import org.apache.spark.internal.LogKeys.{INTERVAL, SESSION_HOLD_INFO}
 import org.apache.spark.sql.classic.SparkSession
@@ -39,6 +39,9 @@ import org.apache.spark.util.ThreadUtils
  */
 class SparkConnectSessionManager extends Logging {
 
+  // Base SparkSession created from the SparkContext, used to create new isolated sessions
+  @volatile private var baseSession: Option[SparkSession] = None
+
   private val sessionStore: ConcurrentMap[SessionKey, SessionHolder] =
     new ConcurrentHashMap[SessionKey, SessionHolder]()
 
@@ -48,6 +51,16 @@ class SparkConnectSessionManager extends Logging {
       .maximumSize(SparkEnv.get.conf.get(CONNECT_SESSION_MANAGER_CLOSED_SESSIONS_TOMBSTONES_SIZE))
       .build[SessionKey, SessionHolderInfo]()
 
+  /**
+   * Initialize the base SparkSession from the provided SparkContext.
+   * This should be called once during SparkConnectService startup.
+   */
+  def initializeBaseSession(sc: SparkContext): Unit = synchronized {
+    if (baseSession.isEmpty) {
+      baseSession = Some(SparkSession.builder().sparkContext(sc).getOrCreate())
+    }
+  }
+
   /** Executor for the periodic maintenance */
   private val scheduledExecutor: AtomicReference[ScheduledExecutorService] =
     new AtomicReference[ScheduledExecutorService]()
@@ -333,13 +346,7 @@ class SparkConnectSessionManager extends Logging {
   }
 
   private def newIsolatedSession(): SparkSession = {
-    val active = SparkSession.active
-    if (active.sparkContext.isStopped) {
-      assert(SparkSession.getDefaultSession.nonEmpty)
-      SparkSession.getDefaultSession.get.newSession()
-    } else {
-      active.newSession()
-    }
+    baseSession.get.newSession()
   }
 
   private def validateSessionCreate(key: SessionKey): Unit = {

sql/connect/server/src/test/scala/org/apache/spark/sql/connect/service/SparkConnectSessionManagerSuite.scala

@@ -32,6 +32,7 @@ class SparkConnectSessionManagerSuite extends SharedSparkSession with BeforeAndA
   override def beforeEach(): Unit = {
     super.beforeEach()
     SparkConnectService.sessionManager.invalidateAllSessions()
+    SparkConnectService.sessionManager.initializeBaseSession(spark.sparkContext)
   }
 
   test("sessionId needs to be an UUID") {
@@ -171,4 +172,46 @@ class SparkConnectSessionManagerSuite extends SharedSparkSession with BeforeAndA
       sessionHolder.getPipelineExecution(graphId).isEmpty,
       "pipeline execution was not removed")
   }
+
+  test("initializeBaseSession initializes base session from SparkContext") {
+    // Create a new session manager to test initialization
+    val sessionManager = new SparkConnectSessionManager()
+
+    // Initialize the base session with the test SparkContext
+    sessionManager.initializeBaseSession(spark.sparkContext)
+
+    // Create an isolated session and verify it was created successfully
+    val key = SessionKey("user", UUID.randomUUID().toString)
+    val sessionHolder = sessionManager.getOrCreateIsolatedSession(key, None)
+
+    // Verify the session was created and is not the same as the active session
+    assert(sessionHolder != null)
+    assert(sessionHolder.session.sessionUUID != spark.sessionUUID)
+
+    // Clean up
+    sessionManager.closeSession(key)
+  }
+
+  test("initializeBaseSession is idempotent") {
+    // Create a new session manager to test initialization
+    val sessionManager = new SparkConnectSessionManager()
+
+    // Initialize the base session multiple times
+    sessionManager.initializeBaseSession(spark.sparkContext)
+    val key1 = SessionKey("user1", UUID.randomUUID().toString)
+    val sessionHolder1 = sessionManager.getOrCreateIsolatedSession(key1, None)
+    val baseSessionUUID1 = sessionHolder1.session.sessionUUID
+
+    // Initialize again - should not change the base session
+    sessionManager.initializeBaseSession(spark.sparkContext)
+    val key2 = SessionKey("user2", UUID.randomUUID().toString)
+    val sessionHolder2 = sessionManager.getOrCreateIsolatedSession(key2, None)
+
+    // Both sessions should be isolated from each other
+    assert(sessionHolder1.session.sessionUUID != sessionHolder2.session.sessionUUID)
+
+    // Clean up
+    sessionManager.closeSession(key1)
+    sessionManager.closeSession(key2)
+  }
 }