YARN-4086. Allow Aggregated Log readers to handle HAR files (rkanter)

hadoop-yarn-project/CHANGES.txt

@@ -428,6 +428,8 @@ Release 2.8.0 - UNRELEASED
     YARN-4121. Fix typos in capacity scheduler documentation.
     (Kai Sasaki via vvasudev)
 
+    YARN-4086. Allow Aggregated Log readers to handle HAR files (rkanter)
+
   OPTIMIZATIONS
 
     YARN-3339. TestDockerContainerExecutor should pull a single image and not

hadoop-yarn-project/hadoop-yarn/hadoop-yarn-client/pom.xml

@@ -134,6 +134,18 @@
 
   <build>
     <plugins>
+      <plugin>
+        <groupId>org.apache.rat</groupId>
+        <artifactId>apache-rat-plugin</artifactId>
+        <configuration>
+          <excludes>
+            <exclude>src/test/resources/application_1440536969523_0001.har/_index</exclude>
+            <exclude>src/test/resources/application_1440536969523_0001.har/part-0</exclude>
+            <exclude>src/test/resources/application_1440536969523_0001.har/_masterindex</exclude>
+            <exclude>src/test/resources/application_1440536969523_0001.har/_SUCCESS</exclude>
+          </excludes>
+        </configuration>
+      </plugin>
       <plugin>
         <groupId>org.apache.hadoop</groupId>
         <artifactId>hadoop-maven-plugins</artifactId>

hadoop-yarn-project/hadoop-yarn/hadoop-yarn-client/src/test/java/org/apache/hadoop/yarn/client/cli/TestLogsCLI.java

@@ -18,6 +18,7 @@
 
 package org.apache.hadoop.yarn.client.cli;
 
+import static org.junit.Assert.assertNotNull;
 import static org.junit.Assert.assertTrue;
 import static org.mockito.Matchers.any;
 import static org.mockito.Mockito.doReturn;
@@ -32,6 +33,7 @@
 import java.io.PrintStream;
 import java.io.PrintWriter;
 import java.io.Writer;
+import java.net.URL;
 import java.util.ArrayList;
 import java.util.Arrays;
 import java.util.HashMap;
@@ -318,6 +320,54 @@ public void testFetchApplictionLogs() throws Exception {
     fs.delete(new Path(rootLogDir), true);
   }
 
+  @Test (timeout = 15000)
+  public void testFetchApplictionLogsHar() throws Exception {
+    String remoteLogRootDir = "target/logs/";
+    Configuration configuration = new Configuration();
+    configuration.setBoolean(YarnConfiguration.LOG_AGGREGATION_ENABLED, true);
+    configuration
+        .set(YarnConfiguration.NM_REMOTE_APP_LOG_DIR, remoteLogRootDir);
+    configuration.setBoolean(YarnConfiguration.YARN_ACL_ENABLE, true);
+    configuration.set(YarnConfiguration.YARN_ADMIN_ACL, "admin");
+    FileSystem fs = FileSystem.get(configuration);
+    UserGroupInformation ugi = UserGroupInformation.getCurrentUser();
+    URL harUrl = ClassLoader.getSystemClassLoader()
+        .getResource("application_1440536969523_0001.har");
+    assertNotNull(harUrl);
+    Path path =
+        new Path(remoteLogRootDir + ugi.getShortUserName()
+            + "/logs/application_1440536969523_0001");
+    if (fs.exists(path)) {
+      fs.delete(path, true);
+    }
+    assertTrue(fs.mkdirs(path));
+    Path harPath = new Path(path, "application_1440536969523_0001.har");
+    fs.copyFromLocalFile(false, new Path(harUrl.toURI()), harPath);
+    assertTrue(fs.exists(harPath));
+
+    YarnClient mockYarnClient =
+        createMockYarnClient(YarnApplicationState.FINISHED);
+    LogsCLI cli = new LogsCLIForTest(mockYarnClient);
+    cli.setConf(configuration);
+    int exitCode = cli.run(new String[]{"-applicationId",
+        "application_1440536969523_0001"});
+    assertTrue(exitCode == 0);
+    String out = sysOutStream.toString();
+    assertTrue(
+        out.contains("container_1440536969523_0001_01_000001 on host1_1111"));
+    assertTrue(out.contains("Hello stderr"));
+    assertTrue(out.contains("Hello stdout"));
+    assertTrue(out.contains("Hello syslog"));
+    assertTrue(
+        out.contains("container_1440536969523_0001_01_000002 on host2_2222"));
+    assertTrue(out.contains("Goodbye stderr"));
+    assertTrue(out.contains("Goodbye stdout"));
+    assertTrue(out.contains("Goodbye syslog"));
+    sysOutStream.reset();
+
+    fs.delete(new Path(remoteLogRootDir), true);
+  }
+
   private static void createContainerLogInLocalDir(Path appLogsDir,
       ContainerId containerId, FileSystem fs, List<String> logTypes) throws Exception {
     Path containerLogsDir = new Path(appLogsDir, containerId.toString());

hadoop-yarn-project/hadoop-yarn/hadoop-yarn-client/src/test/resources/application_1440536969523_0001.har/_index

@@ -0,0 +1,3 @@
+%2F dir 1440540845855+504+rkanter+supergroup 0 0 host1_1111 host2_2222 
+%2Fhost1_1111 file part-0 0 394 1440540845834+420+rkanter+supergroup 
+%2Fhost2_2222 file part-0 394 400 1440540845854+420+rkanter+supergroup 

hadoop-yarn-project/hadoop-yarn/hadoop-yarn-client/src/test/resources/application_1440536969523_0001.har/_masterindex

@@ -0,0 +1,2 @@
+3 
+0 1520266628 0 214 

hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/pom.xml

@@ -215,6 +215,10 @@
             <exclude>src/main/resources/webapps/static/dt-1.9.4/css/demo_table.css</exclude>
             <exclude>src/main/resources/webapps/static/dt-1.9.4/images/Sorting icons.psd</exclude>
             <exclude>src/main/resources/webapps/static/jquery/themes-1.9.1/base/jquery-ui.css</exclude>
+            <exclude>src/test/resources/application_1440536969523_0001.har/_index</exclude>
+            <exclude>src/test/resources/application_1440536969523_0001.har/part-0</exclude>
+            <exclude>src/test/resources/application_1440536969523_0001.har/_masterindex</exclude>
+            <exclude>src/test/resources/application_1440536969523_0001.har/_SUCCESS</exclude>
           </excludes>
         </configuration>
       </plugin>

hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/logaggregation/LogCLIHelpers.java

@@ -31,6 +31,7 @@
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.FileContext;
 import org.apache.hadoop.fs.FileStatus;
+import org.apache.hadoop.fs.HarFs;
 import org.apache.hadoop.fs.Path;
 import org.apache.hadoop.fs.RemoteIterator;
 import org.apache.hadoop.yarn.api.records.ApplicationId;
@@ -61,8 +62,9 @@ public int dumpAContainersLogsForALogType(String appId, String containerId,
         YarnConfiguration.NM_REMOTE_APP_LOG_DIR,
         YarnConfiguration.DEFAULT_NM_REMOTE_APP_LOG_DIR));
     String suffix = LogAggregationUtils.getRemoteNodeLogDirSuffix(getConf());
+    ApplicationId applicationId = ConverterUtils.toApplicationId(appId);
     Path remoteAppLogDir = LogAggregationUtils.getRemoteAppLogDir(
-        remoteRootLogDir, ConverterUtils.toApplicationId(appId), jobOwner,
+        remoteRootLogDir, applicationId, jobOwner,
         suffix);
     RemoteIterator<FileStatus> nodeFiles;
     try {
@@ -80,6 +82,12 @@ public int dumpAContainersLogsForALogType(String appId, String containerId,
     while (nodeFiles.hasNext()) {
       FileStatus thisNodeFile = nodeFiles.next();
       String fileName = thisNodeFile.getPath().getName();
+      if (fileName.equals(applicationId + ".har")) {
+        Path p = new Path("har:///"
+            + thisNodeFile.getPath().toUri().getRawPath());
+        nodeFiles = HarFs.get(p.toUri(), conf).listStatusIterator(p);
+        continue;
+      }
       if (fileName.contains(LogAggregationUtils.getNodeString(nodeId))
           && !fileName.endsWith(LogAggregationUtils.TMP_FILE_SUFFIX)) {
         AggregatedLogFormat.LogReader reader = null;
@@ -207,6 +215,12 @@ public int dumpAllContainersLogs(ApplicationId appId, String appOwner,
     boolean foundAnyLogs = false;
     while (nodeFiles.hasNext()) {
       FileStatus thisNodeFile = nodeFiles.next();
+      if (thisNodeFile.getPath().getName().equals(appId + ".har")) {
+        Path p = new Path("har:///"
+            + thisNodeFile.getPath().toUri().getRawPath());
+        nodeFiles = HarFs.get(p.toUri(), conf).listStatusIterator(p);
+        continue;
+      }
       if (!thisNodeFile.getPath().getName()
         .endsWith(LogAggregationUtils.TMP_FILE_SUFFIX)) {
         AggregatedLogFormat.LogReader reader =

hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/webapp/log/AggregatedLogsBlock.java

@@ -32,6 +32,7 @@
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.FileContext;
 import org.apache.hadoop.fs.FileStatus;
+import org.apache.hadoop.fs.HarFs;
 import org.apache.hadoop.fs.Path;
 import org.apache.hadoop.fs.RemoteIterator;
 import org.apache.hadoop.security.UserGroupInformation;
@@ -120,6 +121,12 @@ protected void render(Block html) {
         AggregatedLogFormat.LogReader reader = null;
         try {
           FileStatus thisNodeFile = nodeFiles.next();
+          if (thisNodeFile.getPath().getName().equals(applicationId + ".har")) {
+            Path p = new Path("har:///"
+                + thisNodeFile.getPath().toUri().getRawPath());
+            nodeFiles = HarFs.get(p.toUri(), conf).listStatusIterator(p);
+            continue;
+          }
           if (!thisNodeFile.getPath().getName()
             .contains(LogAggregationUtils.getNodeString(nodeId))
               || thisNodeFile.getPath().getName()

hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/test/java/org/apache/hadoop/yarn/logaggregation/TestAggregatedLogsBlock.java

@@ -23,6 +23,7 @@
 import java.io.FileWriter;
 import java.io.PrintWriter;
 import java.io.Writer;
+import java.net.URL;
 import java.util.Arrays;
 import java.util.HashMap;
 import java.util.List;
@@ -30,6 +31,7 @@
 
 import javax.servlet.http.HttpServletRequest;
 
+import org.apache.commons.io.FileUtils;
 import org.apache.hadoop.conf.Configuration;
 import org.apache.hadoop.fs.FileUtil;
 import org.apache.hadoop.fs.Path;
@@ -117,7 +119,8 @@ public void testBadLogs() throws Exception {
   }
 
   /**
-   * All ok and the AggregatedLogsBlockFor should aggregate logs and show it.
+   * Reading from logs should succeed and they should be shown in the
+   * AggregatedLogsBlock html.
    * 
    * @throws Exception
    */
@@ -144,8 +147,56 @@ public void testAggregatedLogsBlock() throws Exception {
     assertTrue(out.contains("test log1"));
     assertTrue(out.contains("test log2"));
     assertTrue(out.contains("test log3"));
-
   }
+
+  /**
+   * Reading from logs should succeed (from a HAR archive) and they should be
+   * shown in the AggregatedLogsBlock html.
+   *
+   * @throws Exception
+   */
+  @Test
+  public void testAggregatedLogsBlockHar() throws Exception {
+    FileUtil.fullyDelete(new File("target/logs"));
+    Configuration configuration = getConfiguration();
+
+    URL harUrl = ClassLoader.getSystemClassLoader()
+        .getResource("application_1440536969523_0001.har");
+    assertNotNull(harUrl);
+    String path = "target/logs/admin/logs/application_1440536969523_0001" +
+        "/application_1440536969523_0001.har";
+    FileUtils.copyDirectory(new File(harUrl.getPath()), new File(path));
+
+    AggregatedLogsBlockForTest aggregatedBlock = getAggregatedLogsBlockForTest(
+        configuration, "admin",
+        "container_1440536969523_0001_01_000001", "host1:1111");
+    ByteArrayOutputStream data = new ByteArrayOutputStream();
+    PrintWriter printWriter = new PrintWriter(data);
+    HtmlBlock html = new HtmlBlockForTest();
+    HtmlBlock.Block block = new BlockForTest(html, printWriter, 10, false);
+    aggregatedBlock.render(block);
+
+    block.getWriter().flush();
+    String out = data.toString();
+    assertTrue(out.contains("Hello stderr"));
+    assertTrue(out.contains("Hello stdout"));
+    assertTrue(out.contains("Hello syslog"));
+
+    aggregatedBlock = getAggregatedLogsBlockForTest(
+        configuration, "admin",
+        "container_1440536969523_0001_01_000002", "host2:2222");
+    data = new ByteArrayOutputStream();
+    printWriter = new PrintWriter(data);
+    html = new HtmlBlockForTest();
+    block = new BlockForTest(html, printWriter, 10, false);
+    aggregatedBlock.render(block);
+    block.getWriter().flush();
+    out = data.toString();
+    assertTrue(out.contains("Goodbye stderr"));
+    assertTrue(out.contains("Goodbye stdout"));
+    assertTrue(out.contains("Goodbye syslog"));
+  }
+
   /**
    * Log files was deleted.
    * @throws Exception
@@ -188,14 +239,20 @@ private Configuration getConfiguration() {
 
   private AggregatedLogsBlockForTest getAggregatedLogsBlockForTest(
       Configuration configuration, String user, String containerId) {
+    return getAggregatedLogsBlockForTest(configuration, user, containerId,
+        "localhost:1234");
+  }
+
+  private AggregatedLogsBlockForTest getAggregatedLogsBlockForTest(
+      Configuration configuration, String user, String containerId,
+      String nodeName) {
     HttpServletRequest request = mock(HttpServletRequest.class);
     when(request.getRemoteUser()).thenReturn(user);
     AggregatedLogsBlockForTest aggregatedBlock = new AggregatedLogsBlockForTest(
         configuration);
     aggregatedBlock.setRequest(request);
     aggregatedBlock.moreParams().put(YarnWebParams.CONTAINER_ID, containerId);
-    aggregatedBlock.moreParams().put(YarnWebParams.NM_NODENAME,
+    aggregatedBlock.moreParams().put(YarnWebParams.NM_NODENAME, nodeName);
-        "localhost:1234");
     aggregatedBlock.moreParams().put(YarnWebParams.APP_OWNER, user);
     aggregatedBlock.moreParams().put("start", "");
     aggregatedBlock.moreParams().put("end", "");

hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/test/resources/application_1440536969523_0001.har/_index

@@ -0,0 +1,3 @@
+%2F dir 1440540845855+504+rkanter+supergroup 0 0 host1_1111 host2_2222 
+%2Fhost1_1111 file part-0 0 394 1440540845834+420+rkanter+supergroup 
+%2Fhost2_2222 file part-0 394 400 1440540845854+420+rkanter+supergroup 

hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/test/resources/application_1440536969523_0001.har/_masterindex

@@ -0,0 +1,2 @@
+3 
+0 1520266628 0 214