[hadoop] branch trunk updated: YARN-9419. Log a warning if GPU isolation is enabled but LinuxContainerExecutor is disabled. Contribued by Andras Gyori

[sn...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

snemeth pushed a commit to branch trunk
in repository https://gitbox.apache.org/repos/asf/hadoop.git


The following commit(s) were added to refs/heads/trunk by this push:
     new 44afe11  YARN-9419. Log a warning if GPU isolation is enabled but LinuxContainerExecutor is disabled. Contribued by Andras Gyori
44afe11 is described below

commit 44afe1154dd8ce937470c04a126310989f3dc2cb
Author: Szilard Nemeth <sn...@apache.org>
AuthorDate: Mon Mar 9 16:08:24 2020 +0100

    YARN-9419. Log a warning if GPU isolation is enabled but LinuxContainerExecutor is disabled. Contribued by Andras Gyori
---
 .../resourceplugin/gpu/GpuResourcePlugin.java             | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/resourceplugin/gpu/GpuResourcePlugin.java b/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/resourceplugin/gpu/GpuResourcePlugin.java
index 25ea193..233dd0d 100644
--- a/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/resourceplugin/gpu/GpuResourcePlugin.java
+++ b/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-nodemanager/src/main/java/org/apache/hadoop/yarn/server/nodemanager/containermanager/resourceplugin/gpu/GpuResourcePlugin.java
@@ -20,9 +20,12 @@ package org.apache.hadoop.yarn.server.nodemanager.containermanager.resourceplugi
 
 import java.util.List;
 
+import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.yarn.conf.YarnConfiguration;
 import org.apache.hadoop.yarn.exceptions.YarnException;
+import org.apache.hadoop.yarn.server.nodemanager.ContainerExecutor;
 import org.apache.hadoop.yarn.server.nodemanager.Context;
+import org.apache.hadoop.yarn.server.nodemanager.DefaultContainerExecutor;
 import org.apache.hadoop.yarn.server.nodemanager.containermanager.linux.privileged.PrivilegedOperationExecutor;
 import org.apache.hadoop.yarn.server.nodemanager.containermanager.linux.resources.CGroupsHandler;
 import org.apache.hadoop.yarn.server.nodemanager.containermanager.linux.resources.ResourceHandler;
@@ -59,6 +62,7 @@ public class GpuResourcePlugin implements ResourcePlugin {
 
   @Override
   public void initialize(Context context) throws YarnException {
+    validateExecutorConfig(context.getConf());
     this.gpuDiscoverer.initialize(context.getConf(),
         new NvidiaBinaryHelper());
     this.dockerCommandPlugin =
@@ -66,6 +70,17 @@ public class GpuResourcePlugin implements ResourcePlugin {
             context.getConf());
   }
 
+  private void validateExecutorConfig(Configuration conf) {
+    Class<? extends ContainerExecutor> executorClass = conf.getClass(
+        YarnConfiguration.NM_CONTAINER_EXECUTOR, DefaultContainerExecutor.class,
+        ContainerExecutor.class);
+
+    if (executorClass.equals(DefaultContainerExecutor.class)) {
+      LOG.warn("Using GPU plugin with disabled LinuxContainerExecutor" +
+          " is considered to be unsafe.");
+    }
+  }
+
   @Override
   public ResourceHandler createResourceHandler(
       Context context, CGroupsHandler cGroupsHandler,


---------------------------------------------------------------------
To unsubscribe, e-mail: common-commits-unsubscribe@hadoop.apache.org
For additional commands, e-mail: common-commits-help@hadoop.apache.org