YARN-10333. YarnClient obtain Delegation Token for Log Aggregation Path. Contributed by Prabhu Joseph.

This commit is contained in:
Sunil G 2020-07-09 12:50:25 +05:30
parent dfe60392c9
commit 5dd270e208
3 changed files with 196 additions and 0 deletions

View File

@ -38,6 +38,19 @@
<scope>provided</scope>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-hdfs</artifactId>
<scope>test</scope>
</dependency>
<dependency>
<groupId>org.apache.hadoop</groupId>
<artifactId>hadoop-hdfs</artifactId>
<type>test-jar</type>
<scope>test</scope>
</dependency>
<dependency>
<groupId>com.google.guava</groupId>
<artifactId>guava</artifactId>

View File

@ -30,9 +30,12 @@
import java.util.Set;
import java.util.concurrent.Future;
import org.apache.commons.lang3.StringUtils;
import org.apache.hadoop.classification.InterfaceAudience.Private;
import org.apache.hadoop.classification.InterfaceStability.Unstable;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.DataInputByteBuffer;
import org.apache.hadoop.io.DataOutputBuffer;
import org.apache.hadoop.io.Text;
@ -131,6 +134,8 @@
import org.apache.hadoop.yarn.exceptions.ContainerNotFoundException;
import org.apache.hadoop.yarn.exceptions.YarnException;
import org.apache.hadoop.yarn.exceptions.YarnRuntimeException;
import org.apache.hadoop.yarn.logaggregation.filecontroller.LogAggregationFileController;
import org.apache.hadoop.yarn.logaggregation.filecontroller.LogAggregationFileControllerFactory;
import org.apache.hadoop.yarn.security.AMRMTokenIdentifier;
import org.apache.hadoop.yarn.security.client.TimelineDelegationTokenIdentifier;
import org.apache.hadoop.yarn.util.ConverterUtils;
@ -314,6 +319,16 @@ public YarnClientApplication createApplication()
addTimelineDelegationToken(appContext.getAMContainerSpec());
}
// Automatically add the DT for Log Aggregation path
// This is useful when a separate storage is used for log aggregation
try {
if (isSecurityEnabled()) {
addLogAggregationDelegationToken(appContext.getAMContainerSpec());
}
} catch (Exception e) {
LOG.warn("Failed to obtain delegation token for Log Aggregation Path", e);
}
//TODO: YARN-1763:Handle RM failovers during the submitApplication call.
rmClient.submitApplication(request);
@ -373,6 +388,47 @@ public YarnClientApplication createApplication()
return applicationId;
}
private void addLogAggregationDelegationToken(
ContainerLaunchContext clc) throws YarnException, IOException {
Credentials credentials = new Credentials();
DataInputByteBuffer dibb = new DataInputByteBuffer();
ByteBuffer tokens = clc.getTokens();
if (tokens != null) {
dibb.reset(tokens);
credentials.readTokenStorageStream(dibb);
tokens.rewind();
}
Configuration conf = getConfig();
String masterPrincipal = YarnClientUtils.getRmPrincipal(conf);
if (StringUtils.isEmpty(masterPrincipal)) {
throw new IOException(
"Can't get Master Kerberos principal for use as renewer");
}
LOG.debug("Delegation Token Renewer: " + masterPrincipal);
LogAggregationFileControllerFactory factory =
new LogAggregationFileControllerFactory(conf);
LogAggregationFileController fileController =
factory.getFileControllerForWrite();
Path remoteRootLogDir = fileController.getRemoteRootLogDir();
FileSystem fs = remoteRootLogDir.getFileSystem(conf);
final org.apache.hadoop.security.token.Token<?>[] finalTokens =
fs.addDelegationTokens(masterPrincipal, credentials);
if (finalTokens != null) {
for (org.apache.hadoop.security.token.Token<?> token : finalTokens) {
LOG.info("Added delegation token for log aggregation path "
+ remoteRootLogDir + "; "+token);
}
}
DataOutputBuffer dob = new DataOutputBuffer();
credentials.writeTokenStorageToStream(dob);
tokens = ByteBuffer.wrap(dob.getData(), 0, dob.getLength());
clc.setTokens(tokens);
}
private void addTimelineDelegationToken(
ContainerLaunchContext clc) throws YarnException, IOException {
Credentials credentials = new Credentials();

View File

@ -20,6 +20,12 @@
import org.apache.commons.io.IOUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.FileSystemTestHelper;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.hdfs.HdfsConfiguration;
import org.apache.hadoop.hdfs.MiniDFSCluster;
import org.apache.hadoop.hdfs.security.token.delegation.DelegationTokenIdentifier;
import org.apache.hadoop.io.DataInputByteBuffer;
import org.apache.hadoop.io.DataOutputBuffer;
import org.apache.hadoop.io.Text;
@ -46,12 +52,15 @@
import org.junit.Assert;
import org.junit.Before;
import org.junit.Test;
import org.mockito.invocation.InvocationOnMock;
import org.mockito.stubbing.Answer;
import java.io.IOException;
import java.nio.ByteBuffer;
import java.util.Collection;
import static org.mockito.ArgumentMatchers.any;
import static org.mockito.Mockito.doAnswer;
import static org.mockito.Mockito.mock;
import static org.mockito.Mockito.never;
import static org.mockito.Mockito.spy;
@ -63,6 +72,8 @@
*/
public class TestYarnClientImpl extends ParameterizedSchedulerTestBase {
protected static final String YARN_RM = "yarn-rm@EXAMPLE.COM";
public TestYarnClientImpl(SchedulerType type) throws IOException {
super(type);
}
@ -145,6 +156,122 @@ TimelineClient createTimelineClient() throws IOException, YarnException {
}
}
// Validates if YarnClientImpl automatically adds HDFS Delegation
// token for Log Aggregation Path in a cluster setup with fs.DefaultFS
// set to LocalFileSystem and Log Aggregation Path set to HDFS.
@Test
public void testAutomaitcLogAggregationDelegationToken()
throws Exception {
Configuration conf = getConf();
SecurityUtil.setAuthenticationMethod(
UserGroupInformation.AuthenticationMethod.KERBEROS, conf);
conf.set(YarnConfiguration.RM_PRINCIPAL, YARN_RM);
String remoteRootLogPath = "/tmp/app-logs";
MiniDFSCluster hdfsCluster = null;
try {
// Step 1: Start a MiniDFSCluster for Log Aggregation Path
HdfsConfiguration hdfsConfig = new HdfsConfiguration();
hdfsCluster = new MiniDFSCluster.Builder(hdfsConfig)
.numDataNodes(1).build();
Path remoteRootLogDir = new Path(remoteRootLogPath);
FileSystem fs = hdfsCluster.getFileSystem();
fs.mkdirs(remoteRootLogDir);
conf.set(YarnConfiguration.NM_REMOTE_APP_LOG_DIR,
fs.getFileStatus(remoteRootLogDir).getPath().toString());
// Step 2: Prepare a Mock FileSystem which returns Delegation Token
// when YarnClientImpl invokes
DelegationTokenIdentifier hdfsDT = new DelegationTokenIdentifier(new Text(
"test"), new Text(YARN_RM), null);
final Token<DelegationTokenIdentifier> dToken =
new Token<>(hdfsDT.getBytes(), new byte[0], hdfsDT.getKind(),
new Text());
FileSystem mockFs = mock(FileSystem.class);
doAnswer(new Answer<Token<?>[]>() {
@Override
public Token<?>[] answer(InvocationOnMock invocation) {
Object[] args = invocation.getArguments();
((Credentials) args[1]).addToken(hdfsDT.getKind(), dToken);
return new Token[]{dToken};
}
}).when(mockFs).addDelegationTokens(any(), any());
FileSystemTestHelper.addFileSystemForTesting(fs.getUri(),
hdfsConfig, mockFs);
// Step 3: Prepare a Mock YarnClientImpl
YarnClientImpl client = spy(new YarnClientImpl() {
@Override
protected void serviceStart() {
rmClient = mock(ApplicationClientProtocol.class);
}
@Override
protected void serviceStop() {
}
@Override
public ApplicationReport getApplicationReport(ApplicationId appId) {
ApplicationReport report = mock(ApplicationReport.class);
when(report.getYarnApplicationState())
.thenReturn(YarnApplicationState.RUNNING);
return report;
}
@Override
public boolean isSecurityEnabled() {
return true;
}
});
client.init(conf);
client.start();
// Step 4: Prepare a ApplicationSubmissionContext and submit the app
ApplicationSubmissionContext context =
mock(ApplicationSubmissionContext.class);
ApplicationId applicationId = ApplicationId.newInstance(0, 1);
when(context.getApplicationId()).thenReturn(applicationId);
DataOutputBuffer dob = new DataOutputBuffer();
Credentials credentials = new Credentials();
credentials.writeTokenStorageToStream(dob);
ByteBuffer tokens = ByteBuffer.wrap(dob.getData(), 0, dob.getLength());
ContainerLaunchContext clc = ContainerLaunchContext.newInstance(
null, null, null, null, tokens, null);
when(context.getAMContainerSpec()).thenReturn(clc);
client.submitApplication(context);
// Step 5: Verify automatic addition of HDFS DT for log aggregation path
credentials = new Credentials();
DataInputByteBuffer dibb = new DataInputByteBuffer();
tokens = clc.getTokens();
if (tokens != null) {
dibb.reset(tokens);
credentials.readTokenStorageStream(dibb);
tokens.rewind();
}
Collection<Token<? extends TokenIdentifier>> dTokens =
credentials.getAllTokens();
Assert.assertEquals("Failed to place token for Log Aggregation Path",
1, dTokens.size());
Assert.assertEquals("Wrong Token for Log Aggregation",
hdfsDT.getKind(), dTokens.iterator().next().getKind());
} finally {
if (hdfsCluster != null) {
hdfsCluster.shutdown();
}
}
}
@Test
public void testAutomaticTimelineDelegationTokenLoading()
throws Exception {