org.apache.hadoop.hive.shims.Hadoop23Shims.java Source code

Java tutorial

Introduction

Here is the source code for org.apache.hadoop.hive.shims.Hadoop23Shims.java

Source

/**
 * Licensed to the Apache Software Foundation (ASF) under one
 * or more contributor license agreements.  See the NOTICE file
 * distributed with this work for additional information
 * regarding copyright ownership.  The ASF licenses this file
 * to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance
 * with the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
package org.apache.hadoop.hive.shims;

import java.io.FileNotFoundException;
import java.io.IOException;
import java.lang.reflect.InvocationTargetException;
import java.lang.reflect.Method;
import java.net.InetSocketAddress;
import java.net.MalformedURLException;
import java.net.URI;
import java.nio.ByteBuffer;
import java.security.AccessControlException;
import java.security.NoSuchAlgorithmException;
import java.security.PrivilegedExceptionAction;
import java.util.ArrayList;
import java.util.Comparator;
import java.util.HashSet;
import java.util.Iterator;
import java.util.List;
import java.util.Map;
import java.util.Set;
import java.util.TreeMap;
import javax.security.auth.Subject;

import org.apache.commons.lang.StringUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.crypto.key.KeyProvider;
import org.apache.hadoop.crypto.key.KeyProvider.Options;
import org.apache.hadoop.crypto.key.KeyProviderCryptoExtension;
import org.apache.hadoop.fs.BlockLocation;
import org.apache.hadoop.fs.DefaultFileAccess;
import org.apache.hadoop.fs.FSDataInputStream;
import org.apache.hadoop.fs.FSDataOutputStream;
import org.apache.hadoop.fs.FileStatus;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.LocatedFileStatus;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.fs.PathFilter;
import org.apache.hadoop.fs.ProxyFileSystem;
import org.apache.hadoop.fs.RemoteIterator;
import org.apache.hadoop.fs.TrashPolicy;
import org.apache.hadoop.fs.permission.FsAction;
import org.apache.hadoop.fs.permission.FsPermission;
import org.apache.hadoop.hdfs.DFSClient;
import org.apache.hadoop.hdfs.DistributedFileSystem;
import org.apache.hadoop.hdfs.MiniDFSCluster;
import org.apache.hadoop.hdfs.MiniDFSNNTopology;
import org.apache.hadoop.hdfs.client.HdfsAdmin;
import org.apache.hadoop.hdfs.protocol.DirectoryListing;
import org.apache.hadoop.hdfs.protocol.EncryptionZone;
import org.apache.hadoop.hdfs.protocol.HdfsConstants;
import org.apache.hadoop.hdfs.protocol.HdfsLocatedFileStatus;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.mapred.ClusterStatus;
import org.apache.hadoop.mapred.InputSplit;
import org.apache.hadoop.mapred.JobConf;
import org.apache.hadoop.mapred.MiniMRCluster;
import org.apache.hadoop.mapred.RecordReader;
import org.apache.hadoop.mapred.Reporter;
import org.apache.hadoop.mapred.WebHCatJTShim23;
import org.apache.hadoop.mapred.lib.TotalOrderPartitioner;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.JobContext;
import org.apache.hadoop.mapreduce.JobID;
import org.apache.hadoop.mapreduce.MRJobConfig;
import org.apache.hadoop.mapreduce.OutputFormat;
import org.apache.hadoop.mapreduce.TaskAttemptID;
import org.apache.hadoop.mapreduce.TaskID;
import org.apache.hadoop.mapreduce.TaskType;
import org.apache.hadoop.mapreduce.task.JobContextImpl;
import org.apache.hadoop.mapreduce.task.TaskAttemptContextImpl;
import org.apache.hadoop.net.NetUtils;
import org.apache.hadoop.security.Credentials;
import org.apache.hadoop.security.UserGroupInformation;
import org.apache.hadoop.security.authentication.util.KerberosName;
import org.apache.hadoop.tools.DistCp;
import org.apache.hadoop.tools.DistCpOptions;
import org.apache.hadoop.tools.DistCpOptions.FileAttribute;
import org.apache.hadoop.util.Progressable;
import org.apache.hadoop.yarn.conf.YarnConfiguration;
import org.apache.tez.dag.api.TezConfiguration;
import org.apache.tez.runtime.library.api.TezRuntimeConfiguration;
import org.apache.tez.test.MiniTezCluster;

/**
 * Implemention of shims against Hadoop 0.23.0.
 */
public class Hadoop23Shims extends HadoopShimsSecure {

    HadoopShims.MiniDFSShim cluster = null;
    final boolean storagePolicy;

    public Hadoop23Shims() {
        // in-memory HDFS
        boolean storage = false;
        try {
            Class.forName("org.apache.hadoop.hdfs.protocol.BlockStoragePolicy", false,
                    ShimLoader.class.getClassLoader());
            storage = true;
        } catch (ClassNotFoundException ce) {
        }
        this.storagePolicy = storage;
    }

    @Override
    public HadoopShims.CombineFileInputFormatShim getCombineFileInputFormat() {
        return new CombineFileInputFormatShim() {
            @Override
            public RecordReader getRecordReader(InputSplit split, JobConf job, Reporter reporter)
                    throws IOException {
                throw new IOException("CombineFileInputFormat.getRecordReader not needed.");
            }

            @Override
            protected List<FileStatus> listStatus(JobContext job) throws IOException {
                List<FileStatus> result = super.listStatus(job);
                Iterator<FileStatus> it = result.iterator();
                while (it.hasNext()) {
                    FileStatus stat = it.next();
                    if (!stat.isFile()
                            || (stat.getLen() == 0 && !stat.getPath().toUri().getScheme().equals("nullscan"))) {
                        it.remove();
                    }
                }
                return result;
            }
        };
    }

    @Override
    public String getTaskAttemptLogUrl(JobConf conf, String taskTrackerHttpAddress, String taskAttemptId)
            throws MalformedURLException {
        if (conf.get("mapreduce.framework.name") != null && conf.get("mapreduce.framework.name").equals("yarn")) {
            // if the cluster is running in MR2 mode, return null
            LOG.warn("Can't fetch tasklog: TaskLogServlet is not supported in MR2 mode.");
            return null;
        } else {
            // Was using Hadoop-internal API to get tasklogs, disable until  MAPREDUCE-5857 is fixed.
            LOG.warn("Can't fetch tasklog: TaskLogServlet is not supported in MR1 mode.");
            return null;
        }
    }

    @Override
    public JobTrackerState getJobTrackerState(ClusterStatus clusterStatus) throws Exception {
        switch (clusterStatus.getJobTrackerStatus()) {
        case INITIALIZING:
            return JobTrackerState.INITIALIZING;
        case RUNNING:
            return JobTrackerState.RUNNING;
        default:
            String errorMsg = "Unrecognized JobTracker state: " + clusterStatus.getJobTrackerStatus();
            throw new Exception(errorMsg);
        }
    }

    @Override
    public org.apache.hadoop.mapreduce.TaskAttemptContext newTaskAttemptContext(Configuration conf,
            final Progressable progressable) {
        TaskAttemptID taskAttemptId = TaskAttemptID.forName(conf.get(MRJobConfig.TASK_ATTEMPT_ID));
        if (taskAttemptId == null) {
            // If the caller is not within a mapper/reducer (if reading from the table via CliDriver),
            // then TaskAttemptID.forname() may return NULL. Fall back to using default constructor.
            taskAttemptId = new TaskAttemptID();
        }
        return new TaskAttemptContextImpl(conf, taskAttemptId) {
            @Override
            public void progress() {
                progressable.progress();
            }
        };
    }

    @Override
    public TaskAttemptID newTaskAttemptID(JobID jobId, boolean isMap, int taskId, int id) {
        return new TaskAttemptID(jobId.getJtIdentifier(), jobId.getId(), isMap ? TaskType.MAP : TaskType.REDUCE,
                taskId, id);
    }

    @Override
    public org.apache.hadoop.mapreduce.JobContext newJobContext(Job job) {
        return new JobContextImpl(job.getConfiguration(), job.getJobID());
    }

    @Override
    public boolean isLocalMode(Configuration conf) {
        return "local".equals(conf.get("mapreduce.framework.name"));
    }

    @Override
    public String getJobLauncherRpcAddress(Configuration conf) {
        return conf.get("yarn.resourcemanager.address");
    }

    @Override
    public void setJobLauncherRpcAddress(Configuration conf, String val) {
        if (val.equals("local")) {
            // LocalClientProtocolProvider expects both parameters to be 'local'.
            conf.set("mapreduce.framework.name", val);
            conf.set("mapreduce.jobtracker.address", val);
        } else {
            conf.set("mapreduce.framework.name", "yarn");
            conf.set("yarn.resourcemanager.address", val);
        }
    }

    @Override
    public String getJobLauncherHttpAddress(Configuration conf) {
        return conf.get("yarn.resourcemanager.webapp.address");
    }

    @Override
    public long getDefaultBlockSize(FileSystem fs, Path path) {
        return fs.getDefaultBlockSize(path);
    }

    @Override
    public short getDefaultReplication(FileSystem fs, Path path) {
        return fs.getDefaultReplication(path);
    }

    @Override
    public void setTotalOrderPartitionFile(JobConf jobConf, Path partitionFile) {
        TotalOrderPartitioner.setPartitionFile(jobConf, partitionFile);
    }

    @Override
    public Comparator<LongWritable> getLongComparator() {
        return new Comparator<LongWritable>() {
            @Override
            public int compare(LongWritable o1, LongWritable o2) {
                return o1.compareTo(o2);
            }
        };
    }

    /**
     * Load the fair scheduler queue for given user if available.
     */
    @Override
    public void refreshDefaultQueue(Configuration conf, String userName) throws IOException {
        if (StringUtils.isNotBlank(userName) && isFairScheduler(conf)) {
            ShimLoader.getSchedulerShims().refreshDefaultQueue(conf, userName);
        }
    }

    private boolean isFairScheduler(Configuration conf) {
        return "org.apache.hadoop.yarn.server.resourcemanager.scheduler.fair.FairScheduler"
                .equalsIgnoreCase(conf.get(YarnConfiguration.RM_SCHEDULER));
    }

    /**
     * Returns a shim to wrap MiniMrCluster
     */
    @Override
    public MiniMrShim getMiniMrCluster(Configuration conf, int numberOfTaskTrackers, String nameNode, int numDir)
            throws IOException {
        return new MiniMrShim(conf, numberOfTaskTrackers, nameNode, numDir);
    }

    /**
     * Shim for MiniMrCluster
     */
    public class MiniMrShim implements HadoopShims.MiniMrShim {

        private final MiniMRCluster mr;
        private final Configuration conf;

        public MiniMrShim() {
            mr = null;
            conf = null;
        }

        public MiniMrShim(Configuration conf, int numberOfTaskTrackers, String nameNode, int numDir)
                throws IOException {
            this.conf = conf;

            JobConf jConf = new JobConf(conf);
            jConf.set("yarn.scheduler.capacity.root.queues", "default");
            jConf.set("yarn.scheduler.capacity.root.default.capacity", "100");
            jConf.setInt(MRJobConfig.MAP_MEMORY_MB, 128);
            jConf.setInt(MRJobConfig.REDUCE_MEMORY_MB, 128);
            jConf.setInt(MRJobConfig.MR_AM_VMEM_MB, 128);
            jConf.setInt(YarnConfiguration.YARN_MINICLUSTER_NM_PMEM_MB, 512);
            jConf.setInt(YarnConfiguration.RM_SCHEDULER_MINIMUM_ALLOCATION_MB, 128);
            jConf.setInt(YarnConfiguration.RM_SCHEDULER_MAXIMUM_ALLOCATION_MB, 512);

            mr = new MiniMRCluster(numberOfTaskTrackers, nameNode, numDir, null, null, jConf);
        }

        @Override
        public int getJobTrackerPort() throws UnsupportedOperationException {
            String address = conf.get("yarn.resourcemanager.address");
            address = StringUtils.substringAfterLast(address, ":");

            if (StringUtils.isBlank(address)) {
                throw new IllegalArgumentException("Invalid YARN resource manager port.");
            }

            return Integer.parseInt(address);
        }

        @Override
        public void shutdown() throws IOException {
            mr.shutdown();
        }

        @Override
        public void setupConfiguration(Configuration conf) {
            JobConf jConf = mr.createJobConf();
            for (Map.Entry<String, String> pair : jConf) {
                conf.set(pair.getKey(), pair.getValue());
            }
            conf.setInt(MRJobConfig.MAP_MEMORY_MB, 128);
            conf.setInt(MRJobConfig.REDUCE_MEMORY_MB, 128);
            conf.setInt(MRJobConfig.MR_AM_VMEM_MB, 128);
        }
    }

    @Override
    public HadoopShims.MiniMrShim getLocalMiniTezCluster(Configuration conf, boolean usingLlap) {
        return new MiniTezLocalShim(conf, usingLlap);
    }

    public class MiniTezLocalShim extends Hadoop23Shims.MiniMrShim {
        private final Configuration conf;
        private final boolean isLlap;

        public MiniTezLocalShim(Configuration conf, boolean usingLlap) {
            this.conf = conf;
            this.isLlap = usingLlap;
            setupConfiguration(conf);
        }

        @Override
        public int getJobTrackerPort() throws UnsupportedOperationException {
            throw new UnsupportedOperationException("No JobTracker port for local mode");
        }

        @Override
        public void setupConfiguration(Configuration conf) {
            conf.setBoolean(TezConfiguration.TEZ_LOCAL_MODE, true);

            conf.setBoolean(TezRuntimeConfiguration.TEZ_RUNTIME_OPTIMIZE_LOCAL_FETCH, true);

            conf.setBoolean(TezConfiguration.TEZ_IGNORE_LIB_URIS, true);

            // TODO Force fs to file://, setup staging dir?
            //      conf.set("fs.defaultFS", "file:///");
            //      conf.set(TezConfiguration.TEZ_AM_STAGING_DIR, "/tmp");

            if (!isLlap) { // Conf for non-llap
                conf.setBoolean("hive.llap.io.enabled", false);
            } else { // Conf for llap
                conf.set("hive.llap.execution.mode", "only");
            }
        }

        @Override
        public void shutdown() throws IOException {
            // Nothing to do
        }
    }

    /**
     * Returns a shim to wrap MiniMrTez
     */
    @Override
    public MiniMrShim getMiniTezCluster(Configuration conf, int numberOfTaskTrackers, String nameNode,
            boolean usingLlap) throws IOException {
        return new MiniTezShim(conf, numberOfTaskTrackers, nameNode, usingLlap);
    }

    /**
     * Shim for MiniTezCluster
     */
    public class MiniTezShim extends Hadoop23Shims.MiniMrShim {

        private final MiniTezCluster mr;
        private final Configuration conf;
        private final boolean isLlap;

        public MiniTezShim(Configuration conf, int numberOfTaskTrackers, String nameNode, boolean usingLlap)
                throws IOException {
            mr = new MiniTezCluster("hive", numberOfTaskTrackers);
            conf.setInt(YarnConfiguration.YARN_MINICLUSTER_NM_PMEM_MB, 512);
            conf.setInt(YarnConfiguration.RM_SCHEDULER_MINIMUM_ALLOCATION_MB, 128);
            conf.setInt(YarnConfiguration.RM_SCHEDULER_MAXIMUM_ALLOCATION_MB, 512);
            // Overrides values from the hive/tez-site.
            conf.setInt("hive.tez.container.size", 128);
            conf.setInt(TezConfiguration.TEZ_AM_RESOURCE_MEMORY_MB, 128);
            conf.setInt(TezConfiguration.TEZ_TASK_RESOURCE_MEMORY_MB, 128);
            conf.setInt(TezRuntimeConfiguration.TEZ_RUNTIME_IO_SORT_MB, 24);
            conf.setInt(TezRuntimeConfiguration.TEZ_RUNTIME_UNORDERED_OUTPUT_BUFFER_SIZE_MB, 10);
            conf.setFloat(TezRuntimeConfiguration.TEZ_RUNTIME_SHUFFLE_FETCH_BUFFER_PERCENT, 0.4f);
            conf.set("fs.defaultFS", nameNode);
            conf.set("tez.am.log.level", "DEBUG");
            conf.set(MRJobConfig.MR_AM_STAGING_DIR, "/apps_staging_dir");
            mr.init(conf);
            mr.start();
            this.conf = mr.getConfig();
            this.isLlap = usingLlap;
        }

        @Override
        public int getJobTrackerPort() throws UnsupportedOperationException {
            String address = conf.get("yarn.resourcemanager.address");
            address = StringUtils.substringAfterLast(address, ":");

            if (StringUtils.isBlank(address)) {
                throw new IllegalArgumentException("Invalid YARN resource manager port.");
            }

            return Integer.parseInt(address);
        }

        @Override
        public void shutdown() throws IOException {
            mr.stop();
        }

        @Override
        public void setupConfiguration(Configuration conf) {
            Configuration config = mr.getConfig();
            for (Map.Entry<String, String> pair : config) {
                conf.set(pair.getKey(), pair.getValue());
            }
            // Overrides values from the hive/tez-site.
            conf.setInt("hive.tez.container.size", 128);
            conf.setInt(TezConfiguration.TEZ_AM_RESOURCE_MEMORY_MB, 128);
            conf.setInt(TezConfiguration.TEZ_TASK_RESOURCE_MEMORY_MB, 128);
            conf.setInt(TezRuntimeConfiguration.TEZ_RUNTIME_IO_SORT_MB, 24);
            conf.setInt(TezRuntimeConfiguration.TEZ_RUNTIME_UNORDERED_OUTPUT_BUFFER_SIZE_MB, 10);
            conf.setFloat(TezRuntimeConfiguration.TEZ_RUNTIME_SHUFFLE_FETCH_BUFFER_PERCENT, 0.4f);
            if (isLlap) {
                conf.set("hive.llap.execution.mode", "all");
            }
        }
    }

    private void configureImpersonation(Configuration conf) {
        String user;
        try {
            user = Utils.getUGI().getShortUserName();
        } catch (Exception e) {
            String msg = "Cannot obtain username: " + e;
            throw new IllegalStateException(msg, e);
        }
        conf.set("hadoop.proxyuser." + user + ".groups", "*");
        conf.set("hadoop.proxyuser." + user + ".hosts", "*");
    }

    /**
     * Returns a shim to wrap MiniSparkOnYARNCluster
     */
    @Override
    public MiniMrShim getMiniSparkCluster(Configuration conf, int numberOfTaskTrackers, String nameNode, int numDir)
            throws IOException {
        return new MiniSparkShim(conf, numberOfTaskTrackers, nameNode, numDir);
    }

    /**
     * Shim for MiniSparkOnYARNCluster
     */
    public class MiniSparkShim extends Hadoop23Shims.MiniMrShim {

        private final MiniSparkOnYARNCluster mr;
        private final Configuration conf;

        public MiniSparkShim(Configuration conf, int numberOfTaskTrackers, String nameNode, int numDir)
                throws IOException {
            mr = new MiniSparkOnYARNCluster("sparkOnYarn");
            conf.set("fs.defaultFS", nameNode);
            conf.set("yarn.resourcemanager.scheduler.class",
                    "org.apache.hadoop.yarn.server.resourcemanager.scheduler.fair.FairScheduler");
            // disable resource monitoring, although it should be off by default
            conf.setBoolean(YarnConfiguration.YARN_MINICLUSTER_CONTROL_RESOURCE_MONITORING, false);
            conf.setInt(YarnConfiguration.YARN_MINICLUSTER_NM_PMEM_MB, 2048);
            conf.setInt(YarnConfiguration.RM_SCHEDULER_MINIMUM_ALLOCATION_MB, 512);
            conf.setInt(YarnConfiguration.RM_SCHEDULER_MAXIMUM_ALLOCATION_MB, 2048);
            configureImpersonation(conf);
            mr.init(conf);
            mr.start();
            this.conf = mr.getConfig();
        }

        @Override
        public int getJobTrackerPort() throws UnsupportedOperationException {
            String address = conf.get("yarn.resourcemanager.address");
            address = StringUtils.substringAfterLast(address, ":");

            if (StringUtils.isBlank(address)) {
                throw new IllegalArgumentException("Invalid YARN resource manager port.");
            }

            return Integer.parseInt(address);
        }

        @Override
        public void shutdown() throws IOException {
            mr.stop();
        }

        @Override
        public void setupConfiguration(Configuration conf) {
            Configuration config = mr.getConfig();
            for (Map.Entry<String, String> pair : config) {
                conf.set(pair.getKey(), pair.getValue());
            }
        }
    }

    @Override
    public HadoopShims.MiniDFSShim getMiniDfs(Configuration conf, int numDataNodes, boolean format, String[] racks)
            throws IOException {
        return getMiniDfs(conf, numDataNodes, format, racks, false);
    }

    // Don't move this code to the parent class. There's a binary
    // incompatibility between hadoop 1 and 2 wrt MiniDFSCluster and we
    // need to have two different shim classes even though they are
    // exactly the same.
    @Override
    public HadoopShims.MiniDFSShim getMiniDfs(Configuration conf, int numDataNodes, boolean format, String[] racks,
            boolean isHA) throws IOException {
        configureImpersonation(conf);
        MiniDFSCluster miniDFSCluster;
        if (isHA) {
            MiniDFSNNTopology topo = new MiniDFSNNTopology().addNameservice(new MiniDFSNNTopology.NSConf("minidfs")
                    .addNN(new MiniDFSNNTopology.NNConf("nn1")).addNN(new MiniDFSNNTopology.NNConf("nn2")));
            miniDFSCluster = new MiniDFSCluster.Builder(conf).numDataNodes(numDataNodes).format(format).racks(racks)
                    .nnTopology(topo).build();
            miniDFSCluster.waitActive();
            miniDFSCluster.transitionToActive(0);
        } else {
            miniDFSCluster = new MiniDFSCluster.Builder(conf).numDataNodes(numDataNodes).format(format).racks(racks)
                    .build();
        }

        // Need to set the client's KeyProvider to the NN's for JKS,
        // else the updates do not get flushed properly
        KeyProviderCryptoExtension keyProvider = miniDFSCluster.getNameNode(0).getNamesystem().getProvider();
        if (keyProvider != null) {
            try {
                setKeyProvider(miniDFSCluster.getFileSystem(0).getClient(), keyProvider);
            } catch (Exception err) {
                throw new IOException(err);
            }
        }

        cluster = new MiniDFSShim(miniDFSCluster);
        return cluster;
    }

    private static void setKeyProvider(DFSClient dfsClient, KeyProviderCryptoExtension provider) throws Exception {
        Method setKeyProviderHadoop27Method = null;
        try {
            setKeyProviderHadoop27Method = DFSClient.class.getMethod("setKeyProvider", KeyProvider.class);
        } catch (NoSuchMethodException err) {
            // We can just use setKeyProvider() as it is
        }

        if (setKeyProviderHadoop27Method != null) {
            // Method signature changed in Hadoop 2.7. Cast provider to KeyProvider
            setKeyProviderHadoop27Method.invoke(dfsClient, (KeyProvider) provider);
        } else {
            dfsClient.setKeyProvider(provider);
        }
    }

    /**
     * MiniDFSShim.
     *
     */
    public class MiniDFSShim implements HadoopShims.MiniDFSShim {
        private final MiniDFSCluster cluster;

        public MiniDFSShim(MiniDFSCluster cluster) {
            this.cluster = cluster;
        }

        @Override
        public FileSystem getFileSystem() throws IOException {
            return cluster.getFileSystem(0);
        }

        @Override
        public void shutdown() {
            cluster.shutdown();
        }
    }

    private volatile HCatHadoopShims hcatShimInstance;

    @Override
    public HCatHadoopShims getHCatShim() {
        if (hcatShimInstance == null) {
            hcatShimInstance = new HCatHadoopShims23();
        }
        return hcatShimInstance;
    }

    private final class HCatHadoopShims23 implements HCatHadoopShims {
        @Override
        public TaskID createTaskID() {
            return new TaskID("", 0, TaskType.MAP, 0);
        }

        @Override
        public TaskAttemptID createTaskAttemptID() {
            return new TaskAttemptID("", 0, TaskType.MAP, 0, 0);
        }

        @Override
        public org.apache.hadoop.mapreduce.TaskAttemptContext createTaskAttemptContext(Configuration conf,
                org.apache.hadoop.mapreduce.TaskAttemptID taskId) {
            return new org.apache.hadoop.mapreduce.task.TaskAttemptContextImpl(
                    conf instanceof JobConf ? new JobConf(conf) : conf, taskId);
        }

        @Override
        public org.apache.hadoop.mapred.TaskAttemptContext createTaskAttemptContext(
                org.apache.hadoop.mapred.JobConf conf, org.apache.hadoop.mapred.TaskAttemptID taskId,
                Progressable progressable) {
            org.apache.hadoop.mapred.TaskAttemptContext newContext = null;
            try {
                java.lang.reflect.Constructor<org.apache.hadoop.mapred.TaskAttemptContextImpl> construct = org.apache.hadoop.mapred.TaskAttemptContextImpl.class
                        .getDeclaredConstructor(org.apache.hadoop.mapred.JobConf.class,
                                org.apache.hadoop.mapred.TaskAttemptID.class, Reporter.class);
                construct.setAccessible(true);
                newContext = construct.newInstance(new JobConf(conf), taskId, progressable);
            } catch (Exception e) {
                throw new RuntimeException(e);
            }
            return newContext;
        }

        @Override
        public JobContext createJobContext(Configuration conf, JobID jobId) {
            return new JobContextImpl(conf instanceof JobConf ? new JobConf(conf) : conf, jobId);
        }

        @Override
        public org.apache.hadoop.mapred.JobContext createJobContext(org.apache.hadoop.mapred.JobConf conf,
                org.apache.hadoop.mapreduce.JobID jobId, Progressable progressable) {
            return new org.apache.hadoop.mapred.JobContextImpl(new JobConf(conf), jobId, progressable);
        }

        @Override
        public void commitJob(OutputFormat outputFormat, Job job) throws IOException {
            // Do nothing as this was fixed by MAPREDUCE-1447.
        }

        @Override
        public void abortJob(OutputFormat outputFormat, Job job) throws IOException {
            // Do nothing as this was fixed by MAPREDUCE-1447.
        }

        @Override
        public InetSocketAddress getResourceManagerAddress(Configuration conf) {
            String addr = conf.get("yarn.resourcemanager.address", "localhost:8032");

            return NetUtils.createSocketAddr(addr);
        }

        @Override
        public String getPropertyName(PropertyName name) {
            switch (name) {
            case CACHE_ARCHIVES:
                return MRJobConfig.CACHE_ARCHIVES;
            case CACHE_FILES:
                return MRJobConfig.CACHE_FILES;
            case CACHE_SYMLINK:
                return MRJobConfig.CACHE_SYMLINK;
            case CLASSPATH_ARCHIVES:
                return MRJobConfig.CLASSPATH_ARCHIVES;
            case CLASSPATH_FILES:
                return MRJobConfig.CLASSPATH_FILES;
            }

            return "";
        }

        @Override
        public boolean isFileInHDFS(FileSystem fs, Path path) throws IOException {
            // In case of viewfs we need to lookup where the actual file is to know the filesystem in use.
            // resolvePath is a sure shot way of knowing which file system the file is.
            return "hdfs".equals(fs.resolvePath(path).toUri().getScheme());
        }
    }

    @Override
    public WebHCatJTShim getWebHCatShim(Configuration conf, UserGroupInformation ugi) throws IOException {
        return new WebHCatJTShim23(conf, ugi);//this has state, so can't be cached
    }

    private static final class HdfsFileStatusWithIdImpl implements HdfsFileStatusWithId {
        private final LocatedFileStatus lfs;
        private final long fileId;

        public HdfsFileStatusWithIdImpl(LocatedFileStatus lfs, long fileId) {
            this.lfs = lfs;
            this.fileId = fileId;
        }

        @Override
        public FileStatus getFileStatus() {
            return lfs;
        }

        @Override
        public Long getFileId() {
            return fileId;
        }
    }

    @Override
    public List<HdfsFileStatusWithId> listLocatedHdfsStatus(FileSystem fs, Path p, PathFilter filter)
            throws IOException {
        DistributedFileSystem dfs = ensureDfs(fs);
        DFSClient dfsc = dfs.getClient();
        final String src = p.toUri().getPath();
        DirectoryListing current = dfsc.listPaths(src, org.apache.hadoop.hdfs.protocol.HdfsFileStatus.EMPTY_NAME,
                true);
        if (current == null) { // the directory does not exist
            throw new FileNotFoundException("File " + p + " does not exist.");
        }
        final URI fsUri = fs.getUri();
        List<HdfsFileStatusWithId> result = new ArrayList<HdfsFileStatusWithId>(current.getPartialListing().length);
        while (current != null) {
            org.apache.hadoop.hdfs.protocol.HdfsFileStatus[] hfss = current.getPartialListing();
            for (int i = 0; i < hfss.length; ++i) {
                HdfsLocatedFileStatus next = (HdfsLocatedFileStatus) (hfss[i]);
                if (filter != null) {
                    Path filterPath = next.getFullPath(p).makeQualified(fsUri, null);
                    if (!filter.accept(filterPath))
                        continue;
                }
                LocatedFileStatus lfs = next.makeQualifiedLocated(fsUri, p);
                result.add(new HdfsFileStatusWithIdImpl(lfs, next.getFileId()));
            }
            current = current.hasMore() ? dfsc.listPaths(src, current.getLastName(), true) : null;
        }
        return result;
    }

    private DistributedFileSystem ensureDfs(FileSystem fs) {
        if (!(fs instanceof DistributedFileSystem)) {
            throw new UnsupportedOperationException("Only supported for DFS; got " + fs.getClass());
        }
        DistributedFileSystem dfs = (DistributedFileSystem) fs;
        return dfs;
    }

    @Override
    public BlockLocation[] getLocations(FileSystem fs, FileStatus status) throws IOException {
        if (status instanceof LocatedFileStatus) {
            return ((LocatedFileStatus) status).getBlockLocations();
        } else {
            return fs.getFileBlockLocations(status, 0, status.getLen());
        }
    }

    @Override
    public TreeMap<Long, BlockLocation> getLocationsWithOffset(FileSystem fs, FileStatus status)
            throws IOException {
        TreeMap<Long, BlockLocation> offsetBlockMap = new TreeMap<Long, BlockLocation>();
        BlockLocation[] locations = getLocations(fs, status);
        for (BlockLocation location : locations) {
            offsetBlockMap.put(location.getOffset(), location);
        }
        return offsetBlockMap;
    }

    @Override
    public void hflush(FSDataOutputStream stream) throws IOException {
        stream.hflush();
    }

    class ProxyFileSystem23 extends ProxyFileSystem {
        public ProxyFileSystem23(FileSystem fs) {
            super(fs);
        }

        public ProxyFileSystem23(FileSystem fs, URI uri) {
            super(fs, uri);
        }

        @Override
        public RemoteIterator<LocatedFileStatus> listLocatedStatus(final Path f)
                throws FileNotFoundException, IOException {
            return new RemoteIterator<LocatedFileStatus>() {
                private final RemoteIterator<LocatedFileStatus> stats = ProxyFileSystem23.super.listLocatedStatus(
                        ProxyFileSystem23.super.swizzleParamPath(f));

                @Override
                public boolean hasNext() throws IOException {
                    return stats.hasNext();
                }

                @Override
                public LocatedFileStatus next() throws IOException {
                    LocatedFileStatus result = stats.next();
                    return new LocatedFileStatus(ProxyFileSystem23.super.swizzleFileStatus(result, false),
                            result.getBlockLocations());
                }
            };
        }

        /**
         * Proxy file system also needs to override the access() method behavior.
         * Cannot add Override annotation since FileSystem.access() may not exist in
         * the version of hadoop used to build Hive.
         */
        @Override
        public void access(Path path, FsAction action)
                throws AccessControlException, FileNotFoundException, IOException {
            Path underlyingFsPath = swizzleParamPath(path);
            FileStatus underlyingFsStatus = fs.getFileStatus(underlyingFsPath);
            try {
                if (accessMethod != null) {
                    accessMethod.invoke(fs, underlyingFsPath, action);
                } else {
                    // If the FS has no access() method, we can try DefaultFileAccess ..
                    DefaultFileAccess.checkFileAccess(fs, underlyingFsStatus, action);
                }
            } catch (AccessControlException err) {
                throw err;
            } catch (FileNotFoundException err) {
                throw err;
            } catch (IOException err) {
                throw err;
            } catch (Exception err) {
                throw new RuntimeException(err.getMessage(), err);
            }
        }
    }

    @Override
    public FileSystem createProxyFileSystem(FileSystem fs, URI uri) {
        return new ProxyFileSystem23(fs, uri);
    }

    @Override
    public Configuration getConfiguration(org.apache.hadoop.mapreduce.JobContext context) {
        return context.getConfiguration();
    }

    @Override
    public JobConf getJobConf(org.apache.hadoop.mapred.JobContext context) {
        return context.getJobConf();
    }

    @Override
    public FileSystem getNonCachedFileSystem(URI uri, Configuration conf) throws IOException {
        return FileSystem.newInstance(uri, conf);
    }

    @Override
    public void getMergedCredentials(JobConf jobConf) throws IOException {
        jobConf.getCredentials().mergeAll(UserGroupInformation.getCurrentUser().getCredentials());
    }

    @Override
    public void mergeCredentials(JobConf dest, JobConf src) throws IOException {
        dest.getCredentials().mergeAll(src.getCredentials());
    }

    protected static final Method accessMethod;
    protected static final Method getPasswordMethod;

    static {
        Method m = null;
        try {
            m = FileSystem.class.getMethod("access", Path.class, FsAction.class);
        } catch (NoSuchMethodException err) {
            // This version of Hadoop does not support FileSystem.access().
        }
        accessMethod = m;

        try {
            m = Configuration.class.getMethod("getPassword", String.class);
        } catch (NoSuchMethodException err) {
            // This version of Hadoop does not support getPassword(), just retrieve password from conf.
            m = null;
        }
        getPasswordMethod = m;
    }

    @Override
    public void checkFileAccess(FileSystem fs, FileStatus stat, FsAction action)
            throws IOException, AccessControlException, Exception {
        try {
            if (accessMethod == null) {
                // Have to rely on Hive implementation of filesystem permission checks.
                DefaultFileAccess.checkFileAccess(fs, stat, action);
            } else {
                accessMethod.invoke(fs, stat.getPath(), action);
            }
        } catch (Exception err) {
            throw wrapAccessException(err);
        }
    }

    /**
     * If there is an AccessException buried somewhere in the chain of failures, wrap the original
     * exception in an AccessException. Othewise just return the original exception.
     */
    private static Exception wrapAccessException(Exception err) {
        final int maxDepth = 20;
        Throwable curErr = err;
        for (int idx = 0; curErr != null && idx < maxDepth; ++idx) {
            // fs.permission.AccessControlException removed by HADOOP-11356, but Hive users on older
            // Hadoop versions may still see this exception .. have to reference by name.
            if (curErr instanceof org.apache.hadoop.security.AccessControlException || curErr.getClass().getName()
                    .equals("org.apache.hadoop.fs.permission.AccessControlException")) {
                Exception newErr = new AccessControlException(curErr.getMessage());
                newErr.initCause(err);
                return newErr;
            }
            curErr = curErr.getCause();
        }
        return err;
    }

    @Override
    public String getPassword(Configuration conf, String name) throws IOException {
        if (getPasswordMethod == null) {
            // Just retrieve value from conf
            return conf.get(name);
        } else {
            try {
                char[] pw = (char[]) getPasswordMethod.invoke(conf, name);
                if (pw == null) {
                    return null;
                }
                return new String(pw);
            } catch (Exception err) {
                throw new IOException(err.getMessage(), err);
            }
        }
    }

    @Override
    public boolean supportStickyBit() {
        return true;
    }

    @Override
    public boolean hasStickyBit(FsPermission permission) {
        return permission.getStickyBit();
    }

    @Override
    public boolean supportTrashFeature() {
        return true;
    }

    @Override
    public Path getCurrentTrashPath(Configuration conf, FileSystem fs) {
        TrashPolicy tp = TrashPolicy.getInstance(conf, fs, fs.getHomeDirectory());
        return tp.getCurrentTrashDir();
    }

    /**
     * Returns a shim to wrap KerberosName
     */
    @Override
    public KerberosNameShim getKerberosNameShim(String name) throws IOException {
        return new KerberosNameShim(name);
    }

    @Override
    public boolean isDirectory(FileStatus fileStatus) {
        return fileStatus.isDirectory();
    }

    /**
     * Shim for KerberosName
     */
    public class KerberosNameShim implements HadoopShimsSecure.KerberosNameShim {

        private final KerberosName kerberosName;

        public KerberosNameShim(String name) {
            kerberosName = new KerberosName(name);
        }

        @Override
        public String getDefaultRealm() {
            return kerberosName.getDefaultRealm();
        }

        @Override
        public String getServiceName() {
            return kerberosName.getServiceName();
        }

        @Override
        public String getHostName() {
            return kerberosName.getHostName();
        }

        @Override
        public String getRealm() {
            return kerberosName.getRealm();
        }

        @Override
        public String getShortName() throws IOException {
            return kerberosName.getShortName();
        }
    }

    public static class StoragePolicyShim implements HadoopShims.StoragePolicyShim {

        private final DistributedFileSystem dfs;

        public StoragePolicyShim(DistributedFileSystem fs) {
            this.dfs = fs;
        }

        @Override
        public void setStoragePolicy(Path path, StoragePolicyValue policy) throws IOException {
            switch (policy) {
            case MEMORY: {
                dfs.setStoragePolicy(path, HdfsConstants.MEMORY_STORAGE_POLICY_NAME);
                break;
            }
            case SSD: {
                dfs.setStoragePolicy(path, HdfsConstants.ALLSSD_STORAGE_POLICY_NAME);
                break;
            }
            case DEFAULT: {
                /* do nothing */
                break;
            }
            default:
                throw new IllegalArgumentException("Unknown storage policy " + policy);
            }
        }
    }

    @Override
    public HadoopShims.StoragePolicyShim getStoragePolicyShim(FileSystem fs) {
        if (!storagePolicy) {
            return null;
        }
        try {
            return new StoragePolicyShim((DistributedFileSystem) fs);
        } catch (ClassCastException ce) {
            return null;
        }
    }

    private static final String DISTCP_OPTIONS_PREFIX = "distcp.options.";

    List<String> constructDistCpParams(List<Path> srcPaths, Path dst, Configuration conf) {
        List<String> params = new ArrayList<String>();
        for (Map.Entry<String, String> entry : conf.getPropsWithPrefix(DISTCP_OPTIONS_PREFIX).entrySet()) {
            String distCpOption = entry.getKey();
            String distCpVal = entry.getValue();
            params.add("-" + distCpOption);
            if ((distCpVal != null) && (!distCpVal.isEmpty())) {
                params.add(distCpVal);
            }
        }
        if (params.size() == 0) {
            // if no entries were added via conf, we initiate our defaults
            params.add("-update");
            params.add("-skipcrccheck");
            params.add("-pb");
        }
        for (Path src : srcPaths) {
            params.add(src.toString());
        }
        params.add(dst.toString());
        return params;
    }

    @Override
    public boolean runDistCpAs(List<Path> srcPaths, Path dst, Configuration conf, String doAsUser)
            throws IOException {
        UserGroupInformation proxyUser = UserGroupInformation.createProxyUser(doAsUser,
                UserGroupInformation.getLoginUser());
        try {
            return proxyUser.doAs(new PrivilegedExceptionAction<Boolean>() {
                @Override
                public Boolean run() throws Exception {
                    return runDistCp(srcPaths, dst, conf);
                }
            });
        } catch (InterruptedException e) {
            throw new IOException(e);
        }
    }

    @Override
    public boolean runDistCp(List<Path> srcPaths, Path dst, Configuration conf) throws IOException {
        DistCpOptions options = new DistCpOptions(srcPaths, dst);
        options.setSyncFolder(true);
        options.setSkipCRC(true);
        options.preserve(FileAttribute.BLOCKSIZE);

        // Creates the command-line parameters for distcp
        List<String> params = constructDistCpParams(srcPaths, dst, conf);

        try {
            conf.setBoolean("mapred.mapper.new-api", true);
            DistCp distcp = new DistCp(conf, options);

            // HIVE-13704 states that we should use run() instead of execute() due to a hadoop known issue
            // added by HADOOP-10459
            if (distcp.run(params.toArray(new String[0])) == 0) {
                return true;
            } else {
                return false;
            }
        } catch (Exception e) {
            throw new IOException("Cannot execute DistCp process: " + e, e);
        } finally {
            conf.setBoolean("mapred.mapper.new-api", false);
        }
    }

    private static Boolean hdfsEncryptionSupport;

    public static boolean isHdfsEncryptionSupported() {
        if (hdfsEncryptionSupport == null) {
            Method m = null;

            try {
                m = HdfsAdmin.class.getMethod("getEncryptionZoneForPath", Path.class);
            } catch (NoSuchMethodException e) {
                // This version of Hadoop does not support HdfsAdmin.getEncryptionZoneForPath().
                // Hadoop 2.6.0 introduces this new method.
            }

            hdfsEncryptionSupport = (m != null);
        }

        return hdfsEncryptionSupport;
    }

    public class HdfsEncryptionShim implements HadoopShims.HdfsEncryptionShim {
        private final String HDFS_SECURITY_DEFAULT_CIPHER = "AES/CTR/NoPadding";

        /**
         * Gets information about HDFS encryption zones
         */
        private HdfsAdmin hdfsAdmin = null;

        /**
         * Used to compare encryption key strengths.
         */
        private KeyProvider keyProvider = null;

        private final Configuration conf;

        public HdfsEncryptionShim(URI uri, Configuration conf) throws IOException {
            DistributedFileSystem dfs = (DistributedFileSystem) FileSystem.get(uri, conf);

            this.conf = conf;
            this.keyProvider = dfs.getClient().getKeyProvider();
            this.hdfsAdmin = new HdfsAdmin(uri, conf);
        }

        @Override
        public boolean isPathEncrypted(Path path) throws IOException {
            Path fullPath;
            if (path.isAbsolute()) {
                fullPath = path;
            } else {
                fullPath = path.getFileSystem(conf).makeQualified(path);
            }
            if (!"hdfs".equalsIgnoreCase(path.toUri().getScheme())) {
                return false;
            }
            try {
                return (hdfsAdmin.getEncryptionZoneForPath(fullPath) != null);
            } catch (FileNotFoundException fnfe) {
                LOG.debug("Failed to get EZ for non-existent path: " + fullPath, fnfe);
                return false;
            }
        }

        @Override
        public boolean arePathsOnSameEncryptionZone(Path path1, Path path2) throws IOException {
            return equivalentEncryptionZones(hdfsAdmin.getEncryptionZoneForPath(path1),
                    hdfsAdmin.getEncryptionZoneForPath(path2));
        }

        private boolean equivalentEncryptionZones(EncryptionZone zone1, EncryptionZone zone2) {
            if (zone1 == null && zone2 == null) {
                return true;
            } else if (zone1 == null || zone2 == null) {
                return false;
            }

            return zone1.equals(zone2);
        }

        @Override
        public boolean arePathsOnSameEncryptionZone(Path path1, Path path2,
                HadoopShims.HdfsEncryptionShim encryptionShim2) throws IOException {
            if (!(encryptionShim2 instanceof Hadoop23Shims.HdfsEncryptionShim)) {
                LOG.warn("EncryptionShim for path2 (" + path2 + ") is of unexpected type: "
                        + encryptionShim2.getClass() + ". Assuming path2 is on the same EncryptionZone as path1("
                        + path1 + ").");
                return true;
            }

            return equivalentEncryptionZones(hdfsAdmin.getEncryptionZoneForPath(path1),
                    ((HdfsEncryptionShim) encryptionShim2).hdfsAdmin.getEncryptionZoneForPath(path2));
        }

        @Override
        public int comparePathKeyStrength(Path path1, Path path2) throws IOException {
            EncryptionZone zone1, zone2;

            zone1 = hdfsAdmin.getEncryptionZoneForPath(path1);
            zone2 = hdfsAdmin.getEncryptionZoneForPath(path2);

            if (zone1 == null && zone2 == null) {
                return 0;
            } else if (zone1 == null) {
                return -1;
            } else if (zone2 == null) {
                return 1;
            }

            return compareKeyStrength(zone1.getKeyName(), zone2.getKeyName());
        }

        @Override
        public void createEncryptionZone(Path path, String keyName) throws IOException {
            hdfsAdmin.createEncryptionZone(path, keyName);
        }

        @Override
        public void createKey(String keyName, int bitLength) throws IOException, NoSuchAlgorithmException {

            checkKeyProvider();

            if (keyProvider.getMetadata(keyName) == null) {
                final KeyProvider.Options options = new Options(this.conf);
                options.setCipher(HDFS_SECURITY_DEFAULT_CIPHER);
                options.setBitLength(bitLength);
                keyProvider.createKey(keyName, options);
                keyProvider.flush();
            } else {
                throw new IOException("key '" + keyName + "' already exists");
            }
        }

        @Override
        public void deleteKey(String keyName) throws IOException {
            checkKeyProvider();

            if (keyProvider.getMetadata(keyName) != null) {
                keyProvider.deleteKey(keyName);
                keyProvider.flush();
            } else {
                throw new IOException("key '" + keyName + "' does not exist.");
            }
        }

        @Override
        public List<String> getKeys() throws IOException {
            checkKeyProvider();
            return keyProvider.getKeys();
        }

        private void checkKeyProvider() throws IOException {
            if (keyProvider == null) {
                throw new IOException("HDFS security key provider is not configured on your server.");
            }
        }

        /**
         * Compares two encryption key strengths.
         *
         * @param keyname1 Keyname to compare
         * @param keyname2 Keyname to compare
         * @return 1 if path1 is stronger; 0 if paths are equals; -1 if path1 is weaker.
         * @throws IOException If an error occurred attempting to get key metadata
         */
        private int compareKeyStrength(String keyname1, String keyname2) throws IOException {
            KeyProvider.Metadata meta1, meta2;

            if (keyProvider == null) {
                throw new IOException("HDFS security key provider is not configured on your server.");
            }

            meta1 = keyProvider.getMetadata(keyname1);
            meta2 = keyProvider.getMetadata(keyname2);

            if (meta1.getBitLength() < meta2.getBitLength()) {
                return -1;
            } else if (meta1.getBitLength() == meta2.getBitLength()) {
                return 0;
            } else {
                return 1;
            }
        }
    }

    @Override
    public HadoopShims.HdfsEncryptionShim createHdfsEncryptionShim(FileSystem fs, Configuration conf)
            throws IOException {
        if (isHdfsEncryptionSupported()) {
            URI uri = fs.getUri();
            if ("hdfs".equals(uri.getScheme())) {
                return new HdfsEncryptionShim(uri, conf);
            }
        }

        return new HadoopShims.NoopHdfsEncryptionShim();
    }

    @Override
    public Path getPathWithoutSchemeAndAuthority(Path path) {
        return Path.getPathWithoutSchemeAndAuthority(path);
    }

    @Override
    public int readByteBuffer(FSDataInputStream file, ByteBuffer dest) throws IOException {
        int pos = dest.position();
        int result = file.read(dest);
        if (result > 0) {
            // Ensure this explicitly since versions before 2.7 read doesn't do it.
            dest.position(pos + result);
        }
        return result;
    }

    @Override
    public void addDelegationTokens(FileSystem fs, Credentials cred, String uname) throws IOException {
        // Use method addDelegationTokens instead of getDelegationToken to get all the tokens including KMS.
        fs.addDelegationTokens(uname, cred);
    }

    @Override
    public long getFileId(FileSystem fs, String path) throws IOException {
        return ensureDfs(fs).getClient().getFileInfo(path).getFileId();
    }

    private final static java.lang.reflect.Method getSubjectMethod;
    private final static java.lang.reflect.Constructor<UserGroupInformation> ugiCtor;
    private final static String ugiCloneError;
    static {
        Class<UserGroupInformation> clazz = UserGroupInformation.class;
        java.lang.reflect.Method method = null;
        java.lang.reflect.Constructor<UserGroupInformation> ctor;
        String error = null;
        try {
            method = clazz.getDeclaredMethod("getSubject");
            method.setAccessible(true);
            ctor = clazz.getDeclaredConstructor(Subject.class);
            ctor.setAccessible(true);
        } catch (Throwable t) {
            error = t.getMessage();
            method = null;
            ctor = null;
            LOG.error("Cannot create UGI reflection methods", t);
        }
        getSubjectMethod = method;
        ugiCtor = ctor;
        ugiCloneError = error;
    }

    @Override
    public UserGroupInformation cloneUgi(UserGroupInformation baseUgi) throws IOException {
        // Based on UserGroupInformation::createProxyUser.
        // TODO: use a proper method after we can depend on HADOOP-13081.
        if (getSubjectMethod == null) {
            throw new IOException("The UGI method was not found: " + ugiCloneError);
        }
        try {
            Subject origSubject = (Subject) getSubjectMethod.invoke(baseUgi);

            Subject subject = new Subject(false, origSubject.getPrincipals(),
                    cloneCredentials(origSubject.getPublicCredentials()),
                    cloneCredentials(origSubject.getPrivateCredentials()));
            return ugiCtor.newInstance(subject);
        } catch (InstantiationException | IllegalAccessException | InvocationTargetException e) {
            throw new IOException(e);
        }
    }

    private static Set<Object> cloneCredentials(Set<Object> old) {
        Set<Object> set = new HashSet<>();
        // Make sure Hadoop credentials objects do not reuse the maps.
        for (Object o : old) {
            set.add(o instanceof Credentials ? new Credentials((Credentials) o) : o);
        }
        return set;
    }

}