Repository: spark Updated Branches: refs/heads/master b0a8c16fe -> a920a4369
[SPARK-19038][YARN] Avoid overwriting keytab configuration in yarn-client ## What changes were proposed in this pull request? Because yarn#client will reset the `spark.yarn.keytab` configuration to point to the location in distributed file, so if user still uses the old `SparkConf` to create `SparkSession` with Hive enabled, it will read keytab from the path in distributed cached. This is OK for yarn cluster mode, but in yarn client mode where driver is running out of container, it will be failed to fetch the keytab. So here we should avoid reseting this configuration in the `yarn#client` and only overwriting it for AM, so using `spark.yarn.keytab` could get correct keytab path no matter running in client (keytab in local fs) or cluster (keytab in distributed cache) mode. ## How was this patch tested? Verified in security cluster. Author: jerryshao <[email protected]> Closes #16923 from jerryshao/SPARK-19038. Project: http://git-wip-us.apache.org/repos/asf/spark/repo Commit: http://git-wip-us.apache.org/repos/asf/spark/commit/a920a436 Tree: http://git-wip-us.apache.org/repos/asf/spark/tree/a920a436 Diff: http://git-wip-us.apache.org/repos/asf/spark/diff/a920a436 Branch: refs/heads/master Commit: a920a4369434c84274866a09f61e402232c3b47c Parents: b0a8c16 Author: jerryshao <[email protected]> Authored: Fri Feb 24 09:31:52 2017 -0800 Committer: Marcelo Vanzin <[email protected]> Committed: Fri Feb 24 09:31:52 2017 -0800 ---------------------------------------------------------------------- .../main/scala/org/apache/spark/deploy/yarn/Client.scala | 9 ++++++--- .../org/apache/spark/sql/hive/client/HiveClientImpl.scala | 4 ---- 2 files changed, 6 insertions(+), 7 deletions(-) ---------------------------------------------------------------------- http://git-wip-us.apache.org/repos/asf/spark/blob/a920a436/resource-managers/yarn/src/main/scala/org/apache/spark/deploy/yarn/Client.scala ---------------------------------------------------------------------- diff --git a/resource-managers/yarn/src/main/scala/org/apache/spark/deploy/yarn/Client.scala b/resource-managers/yarn/src/main/scala/org/apache/spark/deploy/yarn/Client.scala index a00234c..fa99cd3 100644 --- a/resource-managers/yarn/src/main/scala/org/apache/spark/deploy/yarn/Client.scala +++ b/resource-managers/yarn/src/main/scala/org/apache/spark/deploy/yarn/Client.scala @@ -100,6 +100,7 @@ private[spark] class Client( private var principal: String = null private var keytab: String = null private var credentials: Credentials = null + private var amKeytabFileName: String = null private val launcherBackend = new LauncherBackend() { override def onStopRequest(): Unit = { @@ -471,7 +472,7 @@ private[spark] class Client( logInfo("To enable the AM to login from keytab, credentials are being copied over to the AM" + " via the YARN Secure Distributed Cache.") val (_, localizedPath) = distribute(keytab, - destName = sparkConf.get(KEYTAB), + destName = Some(amKeytabFileName), appMasterOnly = true) require(localizedPath != null, "Keytab file already distributed.") } @@ -708,6 +709,9 @@ private[spark] class Client( // Save Spark configuration to a file in the archive. val props = new Properties() sparkConf.getAll.foreach { case (k, v) => props.setProperty(k, v) } + // Override spark.yarn.key to point to the location in distributed cache which will be used + // by AM. + Option(amKeytabFileName).foreach { k => props.setProperty(KEYTAB.key, k) } confStream.putNextEntry(new ZipEntry(SPARK_CONF_FILE)) val writer = new OutputStreamWriter(confStream, StandardCharsets.UTF_8) props.store(writer, "Spark configuration.") @@ -995,8 +999,7 @@ private[spark] class Client( val f = new File(keytab) // Generate a file name that can be used for the keytab file, that does not conflict // with any user file. - val keytabFileName = f.getName + "-" + UUID.randomUUID().toString - sparkConf.set(KEYTAB.key, keytabFileName) + amKeytabFileName = f.getName + "-" + UUID.randomUUID().toString sparkConf.set(PRINCIPAL.key, principal) } // Defensive copy of the credentials http://git-wip-us.apache.org/repos/asf/spark/blob/a920a436/sql/hive/src/main/scala/org/apache/spark/sql/hive/client/HiveClientImpl.scala ---------------------------------------------------------------------- diff --git a/sql/hive/src/main/scala/org/apache/spark/sql/hive/client/HiveClientImpl.scala b/sql/hive/src/main/scala/org/apache/spark/sql/hive/client/HiveClientImpl.scala index dc9c3ff..24dfd33 100644 --- a/sql/hive/src/main/scala/org/apache/spark/sql/hive/client/HiveClientImpl.scala +++ b/sql/hive/src/main/scala/org/apache/spark/sql/hive/client/HiveClientImpl.scala @@ -106,10 +106,6 @@ private[hive] class HiveClientImpl( // Set up kerberos credentials for UserGroupInformation.loginUser within // current class loader - // Instead of using the spark conf of the current spark context, a new - // instance of SparkConf is needed for the original value of spark.yarn.keytab - // and spark.yarn.principal set in SparkSubmit, as yarn.Client resets the - // keytab configuration for the link name in distributed cache if (sparkConf.contains("spark.yarn.principal") && sparkConf.contains("spark.yarn.keytab")) { val principalName = sparkConf.get("spark.yarn.principal") val keytabFileName = sparkConf.get("spark.yarn.keytab") --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
