You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@spark.apache.org by GitBox <gi...@apache.org> on 2020/06/09 16:08:03 UTC

[GitHub] [spark] maropu commented on a change in pull request #28751: [SPARK-31926][SQL][test-hive1.2] Fix concurrency issue for ThriftCLIService to getPortNumber

maropu commented on a change in pull request #28751:
URL: https://github.com/apache/spark/pull/28751#discussion_r437361561



##########
File path: sql/hive-thriftserver/src/test/scala/org/apache/spark/sql/hive/thriftserver/SharedThriftServer.scala
##########
@@ -53,11 +55,21 @@ trait SharedThriftServer extends SharedSparkSession {
     }
   }
 
+  protected def jdbcUri: String = if (mode == ServerMode.http) {
+    s"""jdbc:hive2://localhost:$serverPort/
+       |default;
+       |transportMode=http;
+       |httpPath=cliservice
+     """.stripMargin.split("\n").mkString.trim
+  } else {
+    s"""jdbc:hive2://localhost:$serverPort"""
+  }

Review comment:
       nit format:
   ```
   private lazy val jdbcUri = if (mode == ServerMode.http) {
       s"jdbc:hive2://localhost:$serverPort/default;transportMode=http;httpPath=cliservice"
     } else {
       s"jdbc:hive2://localhost:$serverPort"
     }
   ```
   ?

##########
File path: sql/hive-thriftserver/src/test/scala/org/apache/spark/sql/hive/thriftserver/SharedThriftServer.scala
##########
@@ -74,18 +86,31 @@ trait SharedThriftServer extends SharedSparkSession {
     // Set the HIVE_SERVER2_THRIFT_PORT to 0, so it could randomly pick any free port to use.
     // It's much more robust than set a random port generated by ourselves ahead
     sqlContext.setConf(ConfVars.HIVE_SERVER2_THRIFT_PORT.varname, "0")
-    hiveServer2 = HiveThriftServer2.startWithContext(sqlContext)
-    hiveServer2.getServices.asScala.foreach {
-      case t: ThriftCLIService if t.getPortNumber != 0 =>
-        serverPort = t.getPortNumber
-        logInfo(s"Started HiveThriftServer2: port=$serverPort, attempt=$attempt")
-      case _ =>
-    }
+    // Set the HIVE_SERVER2_THRIFT_HTTP_PORT to 0, so it could randomly pick any free port to use.
+    // It's much more robust than set a random port generated by ourselves ahead
+    sqlContext.setConf(ConfVars.HIVE_SERVER2_THRIFT_HTTP_PORT.varname, "0")
+    sqlContext.setConf(ConfVars.HIVE_SERVER2_TRANSPORT_MODE.varname, mode.toString)
+
+    try {
+      hiveServer2 = HiveThriftServer2.startWithContext(sqlContext)
+      hiveServer2.getServices.asScala.foreach {
+        case t: ThriftCLIService if t.getPortNumber != 0 =>
+          serverPort = t.getPortNumber
+          logInfo(s"Started HiveThriftServer2: port=$serverPort, attempt=$attempt")

Review comment:
       Ah, I see. Nice catch.

##########
File path: sql/hive-thriftserver/src/test/scala/org/apache/spark/sql/hive/thriftserver/SharedThriftServer.scala
##########
@@ -71,21 +83,33 @@ trait SharedThriftServer extends SharedSparkSession {
   private def startThriftServer(attempt: Int): Unit = {
     logInfo(s"Trying to start HiveThriftServer2:, attempt=$attempt")
     val sqlContext = spark.newSession().sqlContext
-    // Set the HIVE_SERVER2_THRIFT_PORT to 0, so it could randomly pick any free port to use.
+    // Set the HIVE_SERVER2_THRIFT_PORT and HIVE_SERVER2_THRIFT_HTTP_PORT to 0, so it could
+    // randomly pick any free port to use.
     // It's much more robust than set a random port generated by ourselves ahead
     sqlContext.setConf(ConfVars.HIVE_SERVER2_THRIFT_PORT.varname, "0")
-    hiveServer2 = HiveThriftServer2.startWithContext(sqlContext)
-    hiveServer2.getServices.asScala.foreach {
-      case t: ThriftCLIService if t.getPortNumber != 0 =>
-        serverPort = t.getPortNumber
-        logInfo(s"Started HiveThriftServer2: port=$serverPort, attempt=$attempt")
-      case _ =>
-    }
+    sqlContext.setConf(ConfVars.HIVE_SERVER2_THRIFT_HTTP_PORT.varname, "0")
+    sqlContext.setConf(ConfVars.HIVE_SERVER2_TRANSPORT_MODE.varname, mode.toString)
+
+    try {
+      hiveServer2 = HiveThriftServer2.startWithContext(sqlContext)
+      hiveServer2.getServices.asScala.foreach {
+        case t: ThriftCLIService =>
+          serverPort = t.getPortNumber
+          logInfo(s"Started HiveThriftServer2: port=$serverPort, attempt=$attempt")
+        case _ =>
+      }
 
-    // Wait for thrift server to be ready to serve the query, via executing simple query
-    // till the query succeeds. See SPARK-30345 for more details.
-    eventually(timeout(30.seconds), interval(1.seconds)) {
-      withJdbcStatement { _.execute("SELECT 1") }
+      // Wait for thrift server to be ready to serve the query, via executing simple query
+      // till the query succeeds. See SPARK-30345 for more details.
+      eventually(timeout(30.seconds), interval(1.seconds)) {
+        withJdbcStatement {_.execute("SELECT 1")}

Review comment:
       nit: `withJdbcStatement { _.execute("SELECT 1") }`




----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org