ignite-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From voze...@apache.org
Subject [49/51] [partial] ignite git commit: IGNITE-3916: Initial impl.
Date Fri, 16 Sep 2016 11:21:32 GMT
http://git-wip-us.apache.org/repos/asf/ignite/blob/b7489457/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/fs/v2/IgniteHadoopFileSystem.java
----------------------------------------------------------------------
diff --git a/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/fs/v2/IgniteHadoopFileSystem.java b/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/fs/v2/IgniteHadoopFileSystem.java
new file mode 100644
index 0000000..bd8ed2d
--- /dev/null
+++ b/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/fs/v2/IgniteHadoopFileSystem.java
@@ -0,0 +1,1076 @@
+/*
+ * 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.ignite.hadoop.fs.v2;
+
+import org.apache.commons.logging.Log;
+import org.apache.commons.logging.LogFactory;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.fs.AbstractFileSystem;
+import org.apache.hadoop.fs.BlockLocation;
+import org.apache.hadoop.fs.CreateFlag;
+import org.apache.hadoop.fs.FSDataInputStream;
+import org.apache.hadoop.fs.FSDataOutputStream;
+import org.apache.hadoop.fs.FileChecksum;
+import org.apache.hadoop.fs.FileStatus;
+import org.apache.hadoop.fs.FileSystem;
+import org.apache.hadoop.fs.FsServerDefaults;
+import org.apache.hadoop.fs.FsStatus;
+import org.apache.hadoop.fs.InvalidPathException;
+import org.apache.hadoop.fs.Options;
+import org.apache.hadoop.fs.Path;
+import org.apache.hadoop.fs.permission.FsPermission;
+import org.apache.hadoop.util.DataChecksum;
+import org.apache.hadoop.util.Progressable;
+import org.apache.ignite.IgniteCheckedException;
+import org.apache.ignite.IgniteException;
+import org.apache.ignite.configuration.FileSystemConfiguration;
+import org.apache.ignite.hadoop.fs.HadoopFileSystemFactory;
+import org.apache.ignite.hadoop.fs.IgniteHadoopIgfsSecondaryFileSystem;
+import org.apache.ignite.igfs.IgfsBlockLocation;
+import org.apache.ignite.igfs.IgfsFile;
+import org.apache.ignite.igfs.IgfsMode;
+import org.apache.ignite.igfs.IgfsPath;
+import org.apache.ignite.internal.igfs.common.IgfsLogger;
+import org.apache.ignite.internal.processors.hadoop.igfs.HadoopIgfsEndpoint;
+import org.apache.ignite.internal.processors.hadoop.igfs.HadoopIgfsInputStream;
+import org.apache.ignite.internal.processors.hadoop.igfs.HadoopIgfsOutputStream;
+import org.apache.ignite.internal.processors.hadoop.igfs.HadoopIgfsProxyInputStream;
+import org.apache.ignite.internal.processors.hadoop.igfs.HadoopIgfsProxyOutputStream;
+import org.apache.ignite.internal.processors.hadoop.igfs.HadoopIgfsStreamDelegate;
+import org.apache.ignite.internal.processors.hadoop.igfs.HadoopIgfsWrapper;
+import org.apache.ignite.internal.processors.igfs.IgfsHandshakeResponse;
+import org.apache.ignite.internal.processors.igfs.IgfsModeResolver;
+import org.apache.ignite.internal.processors.igfs.IgfsPaths;
+import org.apache.ignite.internal.processors.igfs.IgfsStatus;
+import org.apache.ignite.internal.processors.igfs.IgfsUtils;
+import org.apache.ignite.internal.util.typedef.F;
+import org.apache.ignite.internal.util.typedef.T2;
+import org.apache.ignite.internal.util.typedef.internal.A;
+import org.apache.ignite.internal.util.typedef.internal.S;
+import org.apache.ignite.internal.util.typedef.internal.U;
+import org.apache.ignite.lifecycle.LifecycleAware;
+import org.jetbrains.annotations.Nullable;
+
+import java.io.BufferedOutputStream;
+import java.io.Closeable;
+import java.io.File;
+import java.io.FileNotFoundException;
+import java.io.IOException;
+import java.io.OutputStream;
+import java.net.URI;
+import java.net.URISyntaxException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collection;
+import java.util.EnumSet;
+import java.util.List;
+import java.util.Map;
+import java.util.concurrent.atomic.AtomicBoolean;
+
+import static org.apache.ignite.configuration.FileSystemConfiguration.DFLT_IGFS_LOG_BATCH_SIZE;
+import static org.apache.ignite.configuration.FileSystemConfiguration.DFLT_IGFS_LOG_DIR;
+import static org.apache.ignite.hadoop.fs.v1.IgniteHadoopFileSystem.getFsHadoopUser;
+import static org.apache.ignite.igfs.IgfsMode.PROXY;
+import static org.apache.ignite.internal.processors.hadoop.fs.HadoopParameters.PARAM_IGFS_COLOCATED_WRITES;
+import static org.apache.ignite.internal.processors.hadoop.fs.HadoopParameters.PARAM_IGFS_LOG_BATCH_SIZE;
+import static org.apache.ignite.internal.processors.hadoop.fs.HadoopParameters.PARAM_IGFS_LOG_DIR;
+import static org.apache.ignite.internal.processors.hadoop.fs.HadoopParameters.PARAM_IGFS_LOG_ENABLED;
+import static org.apache.ignite.internal.processors.hadoop.fs.HadoopParameters.PARAM_IGFS_PREFER_LOCAL_WRITES;
+import static org.apache.ignite.internal.processors.hadoop.fs.HadoopParameters.PARAM_IGFS_SEQ_READS_BEFORE_PREFETCH;
+import static org.apache.ignite.internal.processors.hadoop.igfs.HadoopIgfsUtils.parameter;
+import static org.apache.ignite.internal.processors.igfs.IgfsEx.IGFS_SCHEME;
+
+/**
+ * {@code IGFS} Hadoop 2.x file system driver over file system API. To use
+ * {@code IGFS} as Hadoop file system, you should configure this class
+ * in Hadoop's {@code core-site.xml} as follows:
+ * <pre name="code" class="xml">
+ *  &lt;property&gt;
+ *      &lt;name&gt;fs.default.name&lt;/name&gt;
+ *      &lt;value&gt;igfs://ipc&lt;/value&gt;
+ *  &lt;/property&gt;
+ *
+ *  &lt;property&gt;
+ *      &lt;name&gt;fs.igfs.impl&lt;/name&gt;
+ *      &lt;value&gt;org.apache.ignite.hadoop.fs.v2.IgniteHadoopFileSystem&lt;/value&gt;
+ *  &lt;/property&gt;
+ * </pre>
+ * You should also add Ignite JAR and all libraries to Hadoop classpath. To
+ * do this, add following lines to {@code conf/hadoop-env.sh} script in Hadoop
+ * distribution:
+ * <pre name="code" class="bash">
+ * export IGNITE_HOME=/path/to/Ignite/distribution
+ * export HADOOP_CLASSPATH=$IGNITE_HOME/ignite*.jar
+ *
+ * for f in $IGNITE_HOME/libs/*.jar; do
+ *  export HADOOP_CLASSPATH=$HADOOP_CLASSPATH:$f;
+ * done
+ * </pre>
+ * <h1 class="header">Data vs Clients Nodes</h1>
+ * Hadoop needs to use its FileSystem remotely from client nodes as well as directly on
+ * data nodes. Client nodes are responsible for basic file system operations as well as
+ * accessing data nodes remotely. Usually, client nodes are started together
+ * with {@code job-submitter} or {@code job-scheduler} processes, while data nodes are usually
+ * started together with Hadoop {@code task-tracker} processes.
+ * <p>
+ * For sample client and data node configuration refer to {@code config/hadoop/default-config-client.xml}
+ * and {@code config/hadoop/default-config.xml} configuration files in Ignite installation.
+ */
+public class IgniteHadoopFileSystem extends AbstractFileSystem implements Closeable {
+    /** Logger. */
+    private static final Log LOG = LogFactory.getLog(IgniteHadoopFileSystem.class);
+
+    /** Ensures that close routine is invoked at most once. */
+    private final AtomicBoolean closeGuard = new AtomicBoolean();
+
+    /** Grid remote client. */
+    private HadoopIgfsWrapper rmtClient;
+
+    /** The name of the user this File System created on behalf of. */
+    private final String user;
+
+    /** Working directory. */
+    private IgfsPath workingDir;
+
+    /** URI. */
+    private final URI uri;
+
+    /** Authority. */
+    private String uriAuthority;
+
+    /** Client logger. */
+    private IgfsLogger clientLog;
+
+    /** Server block size. */
+    private long grpBlockSize;
+
+    /** Default replication factor. */
+    private short dfltReplication;
+
+    /** Secondary URI string. */
+    private URI secondaryUri;
+
+    /** Mode resolver. */
+    private IgfsModeResolver modeRslvr;
+
+    /** The secondary file system factory. */
+    private HadoopFileSystemFactory factory;
+
+    /** Whether custom sequential reads before prefetch value is provided. */
+    private boolean seqReadsBeforePrefetchOverride;
+
+    /** Custom-provided sequential reads before prefetch. */
+    private int seqReadsBeforePrefetch;
+
+    /** Flag that controls whether file writes should be colocated on data node. */
+    private boolean colocateFileWrites;
+
+    /** Prefer local writes. */
+    private boolean preferLocFileWrites;
+
+    /**
+     * @param name URI for file system.
+     * @param cfg Configuration.
+     * @throws URISyntaxException if name has invalid syntax.
+     * @throws IOException If initialization failed.
+     */
+    public IgniteHadoopFileSystem(URI name, Configuration cfg) throws URISyntaxException, IOException {
+        super(HadoopIgfsEndpoint.normalize(name), IGFS_SCHEME, false, -1);
+
+        uri = name;
+
+        user = getFsHadoopUser();
+
+        try {
+            initialize(name, cfg);
+        }
+        catch (IOException e) {
+            // Close client if exception occurred.
+            if (rmtClient != null)
+                rmtClient.close(false);
+
+            throw e;
+        }
+
+        workingDir = new IgfsPath("/user/" + user);
+    }
+
+    /** {@inheritDoc} */
+    @Override public void checkPath(Path path) {
+        URI uri = path.toUri();
+
+        if (uri.isAbsolute()) {
+            if (!F.eq(uri.getScheme(), IGFS_SCHEME))
+                throw new InvalidPathException("Wrong path scheme [expected=" + IGFS_SCHEME + ", actual=" +
+                    uri.getAuthority() + ']');
+
+            if (!F.eq(uri.getAuthority(), uriAuthority))
+                throw new InvalidPathException("Wrong path authority [expected=" + uriAuthority + ", actual=" +
+                    uri.getAuthority() + ']');
+        }
+    }
+
+    /**
+     * Public setter that can be used by direct users of FS or Visor.
+     *
+     * @param colocateFileWrites Whether all ongoing file writes should be colocated.
+     */
+    @SuppressWarnings("UnusedDeclaration")
+    public void colocateFileWrites(boolean colocateFileWrites) {
+        this.colocateFileWrites = colocateFileWrites;
+    }
+
+    /**
+     * Enter busy state.
+     *
+     * @throws IOException If file system is stopped.
+     */
+    private void enterBusy() throws IOException {
+        if (closeGuard.get())
+            throw new IOException("File system is stopped.");
+    }
+
+    /**
+     * Leave busy state.
+     */
+    private void leaveBusy() {
+        // No-op.
+    }
+
+    /**
+     * @param name URI passed to constructor.
+     * @param cfg Configuration passed to constructor.
+     * @throws IOException If initialization failed.
+     */
+    @SuppressWarnings("ConstantConditions")
+    private void initialize(URI name, Configuration cfg) throws IOException {
+        enterBusy();
+
+        try {
+            if (rmtClient != null)
+                throw new IOException("File system is already initialized: " + rmtClient);
+
+            A.notNull(name, "name");
+            A.notNull(cfg, "cfg");
+
+            if (!IGFS_SCHEME.equals(name.getScheme()))
+                throw new IOException("Illegal file system URI [expected=" + IGFS_SCHEME +
+                    "://[name]/[optional_path], actual=" + name + ']');
+
+            uriAuthority = name.getAuthority();
+
+            // Override sequential reads before prefetch if needed.
+            seqReadsBeforePrefetch = parameter(cfg, PARAM_IGFS_SEQ_READS_BEFORE_PREFETCH, uriAuthority, 0);
+
+            if (seqReadsBeforePrefetch > 0)
+                seqReadsBeforePrefetchOverride = true;
+
+            // In Ignite replication factor is controlled by data cache affinity.
+            // We use replication factor to force the whole file to be stored on local node.
+            dfltReplication = (short)cfg.getInt("dfs.replication", 3);
+
+            // Get file colocation control flag.
+            colocateFileWrites = parameter(cfg, PARAM_IGFS_COLOCATED_WRITES, uriAuthority, false);
+            preferLocFileWrites = cfg.getBoolean(PARAM_IGFS_PREFER_LOCAL_WRITES, false);
+
+            // Get log directory.
+            String logDirCfg = parameter(cfg, PARAM_IGFS_LOG_DIR, uriAuthority, DFLT_IGFS_LOG_DIR);
+
+            File logDirFile = U.resolveIgnitePath(logDirCfg);
+
+            String logDir = logDirFile != null ? logDirFile.getAbsolutePath() : null;
+
+            rmtClient = new HadoopIgfsWrapper(uriAuthority, logDir, cfg, LOG, user);
+
+            // Handshake.
+            IgfsHandshakeResponse handshake = rmtClient.handshake(logDir);
+
+            grpBlockSize = handshake.blockSize();
+
+            IgfsPaths paths = handshake.secondaryPaths();
+
+            Boolean logEnabled = parameter(cfg, PARAM_IGFS_LOG_ENABLED, uriAuthority, false);
+
+            if (handshake.sampling() != null ? handshake.sampling() : logEnabled) {
+                // Initiate client logger.
+                if (logDir == null)
+                    throw new IOException("Failed to resolve log directory: " + logDirCfg);
+
+                Integer batchSize = parameter(cfg, PARAM_IGFS_LOG_BATCH_SIZE, uriAuthority, DFLT_IGFS_LOG_BATCH_SIZE);
+
+                clientLog = IgfsLogger.logger(uriAuthority, handshake.igfsName(), logDir, batchSize);
+            }
+            else
+                clientLog = IgfsLogger.disabledLogger();
+
+            try {
+                modeRslvr = new IgfsModeResolver(paths.defaultMode(), paths.pathModes());
+            }
+            catch (IgniteCheckedException ice) {
+                throw new IOException(ice);
+            }
+
+            boolean initSecondary = paths.defaultMode() == PROXY;
+
+            if (!initSecondary && paths.pathModes() != null) {
+                for (T2<IgfsPath, IgfsMode> pathMode : paths.pathModes()) {
+                    IgfsMode mode = pathMode.getValue();
+
+                    if (mode == PROXY) {
+                        initSecondary = true;
+
+                        break;
+                    }
+                }
+            }
+
+            if (initSecondary) {
+                try {
+                    factory = (HadoopFileSystemFactory) paths.getPayload(getClass().getClassLoader());
+                }
+                catch (IgniteCheckedException e) {
+                    throw new IOException("Failed to get secondary file system factory.", e);
+                }
+
+                if (factory == null)
+                    throw new IOException("Failed to get secondary file system factory (did you set " +
+                        IgniteHadoopIgfsSecondaryFileSystem.class.getName() + " as \"secondaryFIleSystem\" in " +
+                        FileSystemConfiguration.class.getName() + "?)");
+
+                assert factory != null;
+
+                if (factory instanceof LifecycleAware)
+                    ((LifecycleAware) factory).start();
+
+                try {
+                    FileSystem secFs = factory.get(user);
+
+                    secondaryUri = secFs.getUri();
+
+                    A.ensure(secondaryUri != null, "Secondary file system uri should not be null.");
+                }
+                catch (IOException e) {
+                    throw new IOException("Failed to connect to the secondary file system: " + secondaryUri, e);
+                }
+            }
+        }
+        finally {
+            leaveBusy();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public void close() throws IOException {
+        if (closeGuard.compareAndSet(false, true)) {
+            if (rmtClient == null)
+                return;
+
+            rmtClient.close(false);
+
+            if (clientLog.isLogEnabled())
+                clientLog.close();
+
+            if (factory instanceof LifecycleAware)
+                ((LifecycleAware) factory).stop();
+
+            // Reset initialized resources.
+            rmtClient = null;
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public URI getUri() {
+        return uri;
+    }
+
+    /** {@inheritDoc} */
+    @Override public int getUriDefaultPort() {
+        return -1;
+    }
+
+    /** {@inheritDoc} */
+    @Override public FsServerDefaults getServerDefaults() throws IOException {
+        return new FsServerDefaults(grpBlockSize, (int)grpBlockSize, (int)grpBlockSize, dfltReplication, 64 * 1024,
+            false, 0, DataChecksum.Type.NULL);
+    }
+
+    /** {@inheritDoc} */
+    @Override public boolean setReplication(Path f, short replication) throws IOException {
+        return mode(f) == PROXY && secondaryFileSystem().setReplication(f, replication);
+    }
+
+    /** {@inheritDoc} */
+    @Override public void setTimes(Path f, long mtime, long atime) throws IOException {
+        if (mode(f) == PROXY)
+            secondaryFileSystem().setTimes(f, mtime, atime);
+        else {
+            if (mtime == -1 && atime == -1)
+                return;
+
+            rmtClient.setTimes(convert(f), atime, mtime);
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public FsStatus getFsStatus() throws IOException {
+        IgfsStatus status = rmtClient.fsStatus();
+
+        return new FsStatus(status.spaceTotal(), status.spaceUsed(), status.spaceTotal() - status.spaceUsed());
+    }
+
+    /** {@inheritDoc} */
+    @Override public void setPermission(Path p, FsPermission perm) throws IOException {
+        enterBusy();
+
+        try {
+            A.notNull(p, "p");
+
+            if (mode(p) == PROXY)
+                secondaryFileSystem().setPermission(toSecondary(p), perm);
+            else {
+                if (rmtClient.update(convert(p), permission(perm)) == null)
+                    throw new IOException("Failed to set file permission (file not found?)" +
+                        " [path=" + p + ", perm=" + perm + ']');
+            }
+        }
+        finally {
+            leaveBusy();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public void setOwner(Path p, String usr, String grp) throws IOException {
+        A.notNull(p, "p");
+        A.notNull(usr, "username");
+        A.notNull(grp, "grpName");
+
+        enterBusy();
+
+        try {
+            if (mode(p) == PROXY)
+                secondaryFileSystem().setOwner(toSecondary(p), usr, grp);
+            else if (rmtClient.update(convert(p), F.asMap(IgfsUtils.PROP_USER_NAME, usr,
+                IgfsUtils.PROP_GROUP_NAME, grp)) == null) {
+                throw new IOException("Failed to set file permission (file not found?)" +
+                    " [path=" + p + ", username=" + usr + ", grpName=" + grp + ']');
+            }
+        }
+        finally {
+            leaveBusy();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public FSDataInputStream open(Path f, int bufSize) throws IOException {
+        A.notNull(f, "f");
+
+        enterBusy();
+
+        try {
+            IgfsPath path = convert(f);
+            IgfsMode mode = modeRslvr.resolveMode(path);
+
+            if (mode == PROXY) {
+                FSDataInputStream is = secondaryFileSystem().open(toSecondary(f), bufSize);
+
+                if (clientLog.isLogEnabled()) {
+                    // At this point we do not know file size, so we perform additional request to remote FS to get it.
+                    FileStatus status = secondaryFileSystem().getFileStatus(toSecondary(f));
+
+                    long size = status != null ? status.getLen() : -1;
+
+                    long logId = IgfsLogger.nextId();
+
+                    clientLog.logOpen(logId, path, PROXY, bufSize, size);
+
+                    return new FSDataInputStream(new HadoopIgfsProxyInputStream(is, clientLog, logId));
+                }
+                else
+                    return is;
+            }
+            else {
+                HadoopIgfsStreamDelegate stream = seqReadsBeforePrefetchOverride ?
+                    rmtClient.open(path, seqReadsBeforePrefetch) : rmtClient.open(path);
+
+                long logId = -1;
+
+                if (clientLog.isLogEnabled()) {
+                    logId = IgfsLogger.nextId();
+
+                    clientLog.logOpen(logId, path, mode, bufSize, stream.length());
+                }
+
+                if (LOG.isDebugEnabled())
+                    LOG.debug("Opening input stream [thread=" + Thread.currentThread().getName() + ", path=" + path +
+                        ", bufSize=" + bufSize + ']');
+
+                HadoopIgfsInputStream igfsIn = new HadoopIgfsInputStream(stream, stream.length(),
+                    bufSize, LOG, clientLog, logId);
+
+                if (LOG.isDebugEnabled())
+                    LOG.debug("Opened input stream [path=" + path + ", delegate=" + stream + ']');
+
+                return new FSDataInputStream(igfsIn);
+            }
+        }
+        finally {
+            leaveBusy();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @SuppressWarnings("deprecation")
+    @Override public FSDataOutputStream createInternal(
+        Path f,
+        EnumSet<CreateFlag> flag,
+        FsPermission perm,
+        int bufSize,
+        short replication,
+        long blockSize,
+        Progressable progress,
+        Options.ChecksumOpt checksumOpt,
+        boolean createParent
+    ) throws IOException {
+        A.notNull(f, "f");
+
+        enterBusy();
+
+        boolean overwrite = flag.contains(CreateFlag.OVERWRITE);
+        boolean append = flag.contains(CreateFlag.APPEND);
+        boolean create = flag.contains(CreateFlag.CREATE);
+
+        OutputStream out = null;
+
+        try {
+            IgfsPath path = convert(f);
+            IgfsMode mode = modeRslvr.resolveMode(path);
+
+            if (LOG.isDebugEnabled())
+                LOG.debug("Opening output stream in create [thread=" + Thread.currentThread().getName() + "path=" +
+                    path + ", overwrite=" + overwrite + ", bufSize=" + bufSize + ']');
+
+            if (mode == PROXY) {
+                FSDataOutputStream os = secondaryFileSystem().create(toSecondary(f), perm, flag, bufSize,
+                    replication, blockSize, progress);
+
+                if (clientLog.isLogEnabled()) {
+                    long logId = IgfsLogger.nextId();
+
+                    if (append)
+                        clientLog.logAppend(logId, path, PROXY, bufSize); // Don't have stream ID.
+                    else
+                        clientLog.logCreate(logId, path, PROXY, overwrite, bufSize, replication, blockSize);
+
+                    return new FSDataOutputStream(new HadoopIgfsProxyOutputStream(os, clientLog, logId));
+                }
+                else
+                    return os;
+            }
+            else {
+                Map<String, String> permMap = F.asMap(IgfsUtils.PROP_PERMISSION, toString(perm),
+                    IgfsUtils.PROP_PREFER_LOCAL_WRITES, Boolean.toString(preferLocFileWrites));
+
+                // Create stream and close it in the 'finally' section if any sequential operation failed.
+                HadoopIgfsStreamDelegate stream;
+
+                long logId = -1;
+
+                if (append) {
+                    stream = rmtClient.append(path, create, permMap);
+
+                    if (clientLog.isLogEnabled()) {
+                        logId = IgfsLogger.nextId();
+
+                        clientLog.logAppend(logId, path, mode, bufSize);
+                    }
+
+                    if (LOG.isDebugEnabled())
+                        LOG.debug("Opened output stream in append [path=" + path + ", delegate=" + stream + ']');
+                }
+                else {
+                    stream = rmtClient.create(path, overwrite, colocateFileWrites, replication, blockSize,
+                        permMap);
+
+                    if (clientLog.isLogEnabled()) {
+                        logId = IgfsLogger.nextId();
+
+                        clientLog.logCreate(logId, path, mode, overwrite, bufSize, replication, blockSize);
+                    }
+
+                    if (LOG.isDebugEnabled())
+                        LOG.debug("Opened output stream in create [path=" + path + ", delegate=" + stream + ']');
+                }
+
+                assert stream != null;
+
+                HadoopIgfsOutputStream igfsOut = new HadoopIgfsOutputStream(stream, LOG,
+                    clientLog, logId);
+
+                bufSize = Math.max(64 * 1024, bufSize);
+
+                out = new BufferedOutputStream(igfsOut, bufSize);
+
+                FSDataOutputStream res = new FSDataOutputStream(out, null, 0);
+
+                // Mark stream created successfully.
+                out = null;
+
+                return res;
+            }
+        }
+        finally {
+            // Close if failed during stream creation.
+            if (out != null)
+                U.closeQuiet(out);
+
+            leaveBusy();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public boolean supportsSymlinks() {
+        return false;
+    }
+
+    /** {@inheritDoc} */
+    @Override public void renameInternal(Path src, Path dst) throws IOException {
+        A.notNull(src, "src");
+        A.notNull(dst, "dst");
+
+        enterBusy();
+
+        try {
+            IgfsPath srcPath = convert(src);
+            IgfsPath dstPath = convert(dst);
+
+            IgfsMode srcMode = modeRslvr.resolveMode(srcPath);
+
+            if (clientLog.isLogEnabled())
+                clientLog.logRename(srcPath, srcMode, dstPath);
+
+            if (srcMode == PROXY)
+                secondaryFileSystem().rename(toSecondary(src), toSecondary(dst));
+            else
+                rmtClient.rename(srcPath, dstPath);
+        }
+        finally {
+            leaveBusy();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public boolean delete(Path f, boolean recursive) throws IOException {
+        A.notNull(f, "f");
+
+        enterBusy();
+
+        try {
+            IgfsPath path = convert(f);
+
+            IgfsMode mode = modeRslvr.resolveMode(path);
+
+            if (mode == PROXY) {
+                if (clientLog.isLogEnabled())
+                    clientLog.logDelete(path, PROXY, recursive);
+
+                return secondaryFileSystem().delete(toSecondary(f), recursive);
+            }
+
+            boolean res = rmtClient.delete(path, recursive);
+
+            if (clientLog.isLogEnabled())
+                clientLog.logDelete(path, mode, recursive);
+
+            return res;
+        }
+        finally {
+            leaveBusy();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public void setVerifyChecksum(boolean verifyChecksum) throws IOException {
+        // Checksum has effect for secondary FS only.
+        if (factory != null)
+            secondaryFileSystem().setVerifyChecksum(verifyChecksum);
+    }
+
+    /** {@inheritDoc} */
+    @Override public FileChecksum getFileChecksum(Path f) throws IOException {
+        if (mode(f) == PROXY)
+            return secondaryFileSystem().getFileChecksum(f);
+
+        return null;
+    }
+
+    /** {@inheritDoc} */
+    @Override public FileStatus[] listStatus(Path f) throws IOException {
+        A.notNull(f, "f");
+
+        enterBusy();
+
+        try {
+            IgfsPath path = convert(f);
+            IgfsMode mode = modeRslvr.resolveMode(path);
+
+            if (mode == PROXY) {
+                FileStatus[] arr = secondaryFileSystem().listStatus(toSecondary(f));
+
+                if (arr == null)
+                    throw new FileNotFoundException("File " + f + " does not exist.");
+
+                for (int i = 0; i < arr.length; i++)
+                    arr[i] = toPrimary(arr[i]);
+
+                if (clientLog.isLogEnabled()) {
+                    String[] fileArr = new String[arr.length];
+
+                    for (int i = 0; i < arr.length; i++)
+                        fileArr[i] = arr[i].getPath().toString();
+
+                    clientLog.logListDirectory(path, PROXY, fileArr);
+                }
+
+                return arr;
+            }
+            else {
+                Collection<IgfsFile> list = rmtClient.listFiles(path);
+
+                if (list == null)
+                    throw new FileNotFoundException("File " + f + " does not exist.");
+
+                List<IgfsFile> files = new ArrayList<>(list);
+
+                FileStatus[] arr = new FileStatus[files.size()];
+
+                for (int i = 0; i < arr.length; i++)
+                    arr[i] = convert(files.get(i));
+
+                if (clientLog.isLogEnabled()) {
+                    String[] fileArr = new String[arr.length];
+
+                    for (int i = 0; i < arr.length; i++)
+                        fileArr[i] = arr[i].getPath().toString();
+
+                    clientLog.logListDirectory(path, mode, fileArr);
+                }
+
+                return arr;
+            }
+        }
+        finally {
+            leaveBusy();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public void mkdir(Path f, FsPermission perm, boolean createParent) throws IOException {
+        A.notNull(f, "f");
+
+        enterBusy();
+
+        try {
+            IgfsPath path = convert(f);
+            IgfsMode mode = modeRslvr.resolveMode(path);
+
+            if (mode == PROXY) {
+                if (clientLog.isLogEnabled())
+                    clientLog.logMakeDirectory(path, PROXY);
+
+                secondaryFileSystem().mkdirs(toSecondary(f), perm);
+            }
+            else {
+                rmtClient.mkdirs(path, permission(perm));
+
+                if (clientLog.isLogEnabled())
+                    clientLog.logMakeDirectory(path, mode);
+            }
+        }
+        finally {
+            leaveBusy();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public FileStatus getFileStatus(Path f) throws IOException {
+        A.notNull(f, "f");
+
+        enterBusy();
+
+        try {
+            if (mode(f) == PROXY)
+                return toPrimary(secondaryFileSystem().getFileStatus(toSecondary(f)));
+            else {
+                IgfsFile info = rmtClient.info(convert(f));
+
+                if (info == null)
+                    throw new FileNotFoundException("File not found: " + f);
+
+                return convert(info);
+            }
+        }
+        finally {
+            leaveBusy();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public BlockLocation[] getFileBlockLocations(Path path, long start, long len) throws IOException {
+        A.notNull(path, "path");
+
+        IgfsPath igfsPath = convert(path);
+
+        enterBusy();
+
+        try {
+            if (modeRslvr.resolveMode(igfsPath) == PROXY)
+                return secondaryFileSystem().getFileBlockLocations(path, start, len);
+            else {
+                long now = System.currentTimeMillis();
+
+                List<IgfsBlockLocation> affinity = new ArrayList<>(
+                    rmtClient.affinity(igfsPath, start, len));
+
+                BlockLocation[] arr = new BlockLocation[affinity.size()];
+
+                for (int i = 0; i < arr.length; i++)
+                    arr[i] = convert(affinity.get(i));
+
+                if (LOG.isDebugEnabled())
+                    LOG.debug("Fetched file locations [path=" + path + ", fetchTime=" +
+                        (System.currentTimeMillis() - now) + ", locations=" + Arrays.asList(arr) + ']');
+
+                return arr;
+            }
+        }
+        finally {
+            leaveBusy();
+        }
+    }
+
+    /**
+     * Resolve path mode.
+     *
+     * @param path HDFS path.
+     * @return Path mode.
+     */
+    public IgfsMode mode(Path path) {
+        return modeRslvr.resolveMode(convert(path));
+    }
+
+    /**
+     * Convert the given path to path acceptable by the primary file system.
+     *
+     * @param path Path.
+     * @return Primary file system path.
+     */
+    private Path toPrimary(Path path) {
+        return convertPath(path, getUri());
+    }
+
+    /**
+     * Convert the given path to path acceptable by the secondary file system.
+     *
+     * @param path Path.
+     * @return Secondary file system path.
+     */
+    private Path toSecondary(Path path) {
+        assert factory != null;
+        assert secondaryUri != null;
+
+        return convertPath(path, secondaryUri);
+    }
+
+    /**
+     * Convert path using the given new URI.
+     *
+     * @param path Old path.
+     * @param newUri New URI.
+     * @return New path.
+     */
+    private Path convertPath(Path path, URI newUri) {
+        assert newUri != null;
+
+        if (path != null) {
+            URI pathUri = path.toUri();
+
+            try {
+                return new Path(new URI(pathUri.getScheme() != null ? newUri.getScheme() : null,
+                    pathUri.getAuthority() != null ? newUri.getAuthority() : null, pathUri.getPath(), null, null));
+            }
+            catch (URISyntaxException e) {
+                throw new IgniteException("Failed to construct secondary file system path from the primary file " +
+                    "system path: " + path, e);
+            }
+        }
+        else
+            return null;
+    }
+
+    /**
+     * Convert a file status obtained from the secondary file system to a status of the primary file system.
+     *
+     * @param status Secondary file system status.
+     * @return Primary file system status.
+     */
+    private FileStatus toPrimary(FileStatus status) {
+        return status != null ? new FileStatus(status.getLen(), status.isDirectory(), status.getReplication(),
+            status.getBlockSize(), status.getModificationTime(), status.getAccessTime(), status.getPermission(),
+            status.getOwner(), status.getGroup(), toPrimary(status.getPath())) : null;
+    }
+
+    /**
+     * Convert IGFS path into Hadoop path.
+     *
+     * @param path IGFS path.
+     * @return Hadoop path.
+     */
+    private Path convert(IgfsPath path) {
+        return new Path(IGFS_SCHEME, uriAuthority, path.toString());
+    }
+
+    /**
+     * Convert Hadoop path into IGFS path.
+     *
+     * @param path Hadoop path.
+     * @return IGFS path.
+     */
+    @Nullable private IgfsPath convert(Path path) {
+        if (path == null)
+            return null;
+
+        return path.isAbsolute() ? new IgfsPath(path.toUri().getPath()) :
+            new IgfsPath(workingDir, path.toUri().getPath());
+    }
+
+    /**
+     * Convert IGFS affinity block location into Hadoop affinity block location.
+     *
+     * @param block IGFS affinity block location.
+     * @return Hadoop affinity block location.
+     */
+    private BlockLocation convert(IgfsBlockLocation block) {
+        Collection<String> names = block.names();
+        Collection<String> hosts = block.hosts();
+
+        return new BlockLocation(
+            names.toArray(new String[names.size()]) /* hostname:portNumber of data nodes */,
+            hosts.toArray(new String[hosts.size()]) /* hostnames of data nodes */,
+            block.start(), block.length()
+        ) {
+            @Override public String toString() {
+                try {
+                    return "BlockLocation [offset=" + getOffset() + ", length=" + getLength() +
+                        ", hosts=" + Arrays.asList(getHosts()) + ", names=" + Arrays.asList(getNames()) + ']';
+                }
+                catch (IOException e) {
+                    throw new RuntimeException(e);
+                }
+            }
+        };
+    }
+
+    /**
+     * Convert IGFS file information into Hadoop file status.
+     *
+     * @param file IGFS file information.
+     * @return Hadoop file status.
+     */
+    private FileStatus convert(IgfsFile file) {
+        return new FileStatus(
+            file.length(),
+            file.isDirectory(),
+            dfltReplication,
+            file.groupBlockSize(),
+            file.modificationTime(),
+            file.accessTime(),
+            permission(file),
+            file.property(IgfsUtils.PROP_USER_NAME, user),
+            file.property(IgfsUtils.PROP_GROUP_NAME, "users"),
+            convert(file.path())) {
+            @Override public String toString() {
+                return "FileStatus [path=" + getPath() + ", isDir=" + isDirectory() + ", len=" + getLen() + "]";
+            }
+        };
+    }
+
+    /**
+     * Convert Hadoop permission into IGFS file attribute.
+     *
+     * @param perm Hadoop permission.
+     * @return IGFS attributes.
+     */
+    private Map<String, String> permission(FsPermission perm) {
+        if (perm == null)
+            perm = FsPermission.getDefault();
+
+        return F.asMap(IgfsUtils.PROP_PERMISSION, toString(perm));
+    }
+
+    /**
+     * @param perm Permission.
+     * @return String.
+     */
+    private static String toString(FsPermission perm) {
+        return String.format("%04o", perm.toShort());
+    }
+
+    /**
+     * Convert IGFS file attributes into Hadoop permission.
+     *
+     * @param file File info.
+     * @return Hadoop permission.
+     */
+    private FsPermission permission(IgfsFile file) {
+        String perm = file.property(IgfsUtils.PROP_PERMISSION, null);
+
+        if (perm == null)
+            return FsPermission.getDefault();
+
+        try {
+            return new FsPermission((short)Integer.parseInt(perm, 8));
+        }
+        catch (NumberFormatException ignore) {
+            return FsPermission.getDefault();
+        }
+    }
+
+    /** {@inheritDoc} */
+    @Override public String toString() {
+        return S.toString(IgniteHadoopFileSystem.class, this);
+    }
+
+    /**
+     * Returns the user name this File System is created on behalf of.
+     * @return the user name
+     */
+    public String user() {
+        return user;
+    }
+
+    /**
+     * Gets cached or creates a {@link FileSystem}.
+     *
+     * @return The secondary file system.
+     */
+    private FileSystem secondaryFileSystem() throws IOException{
+        assert factory != null;
+
+        return factory.get(user);
+    }
+}
\ No newline at end of file

http://git-wip-us.apache.org/repos/asf/ignite/blob/b7489457/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/fs/v2/package-info.java
----------------------------------------------------------------------
diff --git a/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/fs/v2/package-info.java b/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/fs/v2/package-info.java
new file mode 100644
index 0000000..d8e70d1
--- /dev/null
+++ b/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/fs/v2/package-info.java
@@ -0,0 +1,22 @@
+/*
+ * 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 description. -->
+ * Contains Ignite Hadoop 2.x <code>FileSystem</code> implementation.
+ */
+package org.apache.ignite.hadoop.fs.v2;
\ No newline at end of file

http://git-wip-us.apache.org/repos/asf/ignite/blob/b7489457/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/mapreduce/IgniteHadoopClientProtocolProvider.java
----------------------------------------------------------------------
diff --git a/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/mapreduce/IgniteHadoopClientProtocolProvider.java b/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/mapreduce/IgniteHadoopClientProtocolProvider.java
new file mode 100644
index 0000000..583af35
--- /dev/null
+++ b/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/mapreduce/IgniteHadoopClientProtocolProvider.java
@@ -0,0 +1,144 @@
+/*
+ * 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.ignite.hadoop.mapreduce;
+
+import java.io.IOException;
+import java.net.InetSocketAddress;
+import java.util.Collections;
+import java.util.concurrent.ConcurrentHashMap;
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.mapreduce.MRConfig;
+import org.apache.hadoop.mapreduce.protocol.ClientProtocol;
+import org.apache.hadoop.mapreduce.protocol.ClientProtocolProvider;
+import org.apache.ignite.IgniteCheckedException;
+import org.apache.ignite.internal.IgniteInternalFuture;
+import org.apache.ignite.internal.client.GridClient;
+import org.apache.ignite.internal.client.GridClientConfiguration;
+import org.apache.ignite.internal.client.GridClientException;
+import org.apache.ignite.internal.client.GridClientFactory;
+import org.apache.ignite.internal.client.marshaller.jdk.GridClientJdkMarshaller;
+import org.apache.ignite.internal.processors.hadoop.proto.HadoopClientProtocol;
+import org.apache.ignite.internal.util.future.GridFutureAdapter;
+import org.apache.ignite.internal.util.typedef.F;
+
+import static org.apache.ignite.internal.client.GridClientProtocol.TCP;
+
+
+/**
+ * Ignite Hadoop client protocol provider.
+ */
+public class IgniteHadoopClientProtocolProvider extends ClientProtocolProvider {
+    /** Framework name used in configuration. */
+    public static final String FRAMEWORK_NAME = "ignite";
+
+    /** Clients. */
+    private static final ConcurrentHashMap<String, IgniteInternalFuture<GridClient>> cliMap = new ConcurrentHashMap<>();
+
+    /** {@inheritDoc} */
+    @Override public ClientProtocol create(Configuration conf) throws IOException {
+        if (FRAMEWORK_NAME.equals(conf.get(MRConfig.FRAMEWORK_NAME))) {
+            String addr = conf.get(MRConfig.MASTER_ADDRESS);
+
+            if (F.isEmpty(addr))
+                throw new IOException("Failed to create client protocol because server address is not specified (is " +
+                    MRConfig.MASTER_ADDRESS + " property set?).");
+
+            if (F.eq(addr, "local"))
+                throw new IOException("Local execution mode is not supported, please point " +
+                    MRConfig.MASTER_ADDRESS + " to real Ignite node.");
+
+            return createProtocol(addr, conf);
+        }
+
+        return null;
+    }
+
+    /** {@inheritDoc} */
+    @Override public ClientProtocol create(InetSocketAddress addr, Configuration conf) throws IOException {
+        if (FRAMEWORK_NAME.equals(conf.get(MRConfig.FRAMEWORK_NAME)))
+            return createProtocol(addr.getHostString() + ":" + addr.getPort(), conf);
+
+        return null;
+    }
+
+    /** {@inheritDoc} */
+    @Override public void close(ClientProtocol cliProto) throws IOException {
+        // No-op.
+    }
+
+    /**
+     * Internal protocol creation routine.
+     *
+     * @param addr Address.
+     * @param conf Configuration.
+     * @return Client protocol.
+     * @throws IOException If failed.
+     */
+    private static ClientProtocol createProtocol(String addr, Configuration conf) throws IOException {
+        return new HadoopClientProtocol(conf, client(addr));
+    }
+
+    /**
+     * Create client.
+     *
+     * @param addr Endpoint address.
+     * @return Client.
+     * @throws IOException If failed.
+     */
+    private static GridClient client(String addr) throws IOException {
+        try {
+            IgniteInternalFuture<GridClient> fut = cliMap.get(addr);
+
+            if (fut == null) {
+                GridFutureAdapter<GridClient> fut0 = new GridFutureAdapter<>();
+
+                IgniteInternalFuture<GridClient> oldFut = cliMap.putIfAbsent(addr, fut0);
+
+                if (oldFut != null)
+                    return oldFut.get();
+                else {
+                    GridClientConfiguration cliCfg = new GridClientConfiguration();
+
+                    cliCfg.setProtocol(TCP);
+                    cliCfg.setServers(Collections.singletonList(addr));
+                    cliCfg.setMarshaller(new GridClientJdkMarshaller());
+                    cliCfg.setMaxConnectionIdleTime(24 * 60 * 60 * 1000L); // 1 day.
+                    cliCfg.setDaemon(true);
+
+                    try {
+                        GridClient cli = GridClientFactory.start(cliCfg);
+
+                        fut0.onDone(cli);
+
+                        return cli;
+                    }
+                    catch (GridClientException e) {
+                        fut0.onDone(e);
+
+                        throw new IOException("Failed to establish connection with Ignite node: " + addr, e);
+                    }
+                }
+            }
+            else
+                return fut.get();
+        }
+        catch (IgniteCheckedException e) {
+            throw new IOException("Failed to establish connection with Ignite node: " + addr, e);
+        }
+    }
+}
\ No newline at end of file

http://git-wip-us.apache.org/repos/asf/ignite/blob/b7489457/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/mapreduce/IgniteHadoopMapReducePlanner.java
----------------------------------------------------------------------
diff --git a/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/mapreduce/IgniteHadoopMapReducePlanner.java b/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/mapreduce/IgniteHadoopMapReducePlanner.java
new file mode 100644
index 0000000..d4a44fa
--- /dev/null
+++ b/modules/hadoop-impl/src/main/java/org/apache/ignite/hadoop/mapreduce/IgniteHadoopMapReducePlanner.java
@@ -0,0 +1,416 @@
+/*
+ * 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.ignite.hadoop.mapreduce;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Collections;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.List;
+import java.util.ListIterator;
+import java.util.Map;
+import java.util.UUID;
+
+import org.apache.ignite.IgniteCheckedException;
+import org.apache.ignite.IgniteException;
+import org.apache.ignite.cluster.ClusterNode;
+import org.apache.ignite.igfs.IgfsBlockLocation;
+import org.apache.ignite.igfs.IgfsPath;
+import org.apache.ignite.internal.IgniteEx;
+import org.apache.ignite.internal.processors.hadoop.HadoopFileBlock;
+import org.apache.ignite.internal.processors.hadoop.HadoopInputSplit;
+import org.apache.ignite.internal.processors.hadoop.HadoopJob;
+import org.apache.ignite.internal.processors.hadoop.HadoopMapReducePlan;
+import org.apache.ignite.internal.processors.hadoop.igfs.HadoopIgfsEndpoint;
+import org.apache.ignite.internal.processors.hadoop.planner.HadoopDefaultMapReducePlan;
+import org.apache.ignite.internal.processors.hadoop.planner.HadoopAbstractMapReducePlanner;
+import org.apache.ignite.internal.processors.igfs.IgfsEx;
+import org.apache.ignite.internal.util.typedef.F;
+import org.jetbrains.annotations.NotNull;
+import org.jetbrains.annotations.Nullable;
+
+import static org.apache.ignite.IgniteFileSystem.IGFS_SCHEME;
+
+/**
+ * Default map-reduce planner implementation.
+ */
+public class IgniteHadoopMapReducePlanner extends HadoopAbstractMapReducePlanner {
+    /** {@inheritDoc} */
+    @Override public HadoopMapReducePlan preparePlan(HadoopJob job, Collection<ClusterNode> top,
+        @Nullable HadoopMapReducePlan oldPlan) throws IgniteCheckedException {
+        // Convert collection of topology nodes to collection of topology node IDs.
+        Collection<UUID> topIds = new HashSet<>(top.size(), 1.0f);
+
+        for (ClusterNode topNode : top)
+            topIds.add(topNode.id());
+
+        Map<UUID, Collection<HadoopInputSplit>> mappers = mappers(top, topIds, job.input());
+
+        int rdcCnt = job.info().reducers();
+
+        if (rdcCnt < 0)
+            throw new IgniteCheckedException("Number of reducers must be non-negative, actual: " + rdcCnt);
+
+        Map<UUID, int[]> reducers = reducers(top, mappers, rdcCnt);
+
+        return new HadoopDefaultMapReducePlan(mappers, reducers);
+    }
+
+    /**
+     * Create plan for mappers.
+     *
+     * @param top Topology nodes.
+     * @param topIds Topology node IDs.
+     * @param splits Splits.
+     * @return Mappers map.
+     * @throws IgniteCheckedException If failed.
+     */
+    private Map<UUID, Collection<HadoopInputSplit>> mappers(Collection<ClusterNode> top, Collection<UUID> topIds,
+        Iterable<HadoopInputSplit> splits) throws IgniteCheckedException {
+        Map<UUID, Collection<HadoopInputSplit>> mappers = new HashMap<>();
+
+        Map<String, Collection<UUID>> nodes = groupByHost(top);
+
+        Map<UUID, Integer> nodeLoads = new HashMap<>(top.size(), 1.0f); // Track node load.
+
+        for (UUID nodeId : topIds)
+            nodeLoads.put(nodeId, 0);
+
+        for (HadoopInputSplit split : splits) {
+            UUID nodeId = nodeForSplit(split, topIds, nodes, nodeLoads);
+
+            if (log.isDebugEnabled())
+                log.debug("Mapped split to node [split=" + split + ", nodeId=" + nodeId + ']');
+
+            Collection<HadoopInputSplit> nodeSplits = mappers.get(nodeId);
+
+            if (nodeSplits == null) {
+                nodeSplits = new ArrayList<>();
+
+                mappers.put(nodeId, nodeSplits);
+            }
+
+            nodeSplits.add(split);
+
+            // Updated node load.
+            nodeLoads.put(nodeId, nodeLoads.get(nodeId) + 1);
+        }
+
+        return mappers;
+    }
+
+    /**
+     * Determine the best node for this split.
+     *
+     * @param split Split.
+     * @param topIds Topology node IDs.
+     * @param nodes Nodes.
+     * @param nodeLoads Node load tracker.
+     * @return Node ID.
+     */
+    @SuppressWarnings("unchecked")
+    private UUID nodeForSplit(HadoopInputSplit split, Collection<UUID> topIds, Map<String, Collection<UUID>> nodes,
+        Map<UUID, Integer> nodeLoads) throws IgniteCheckedException {
+        if (split instanceof HadoopFileBlock) {
+            HadoopFileBlock split0 = (HadoopFileBlock)split;
+
+            if (IGFS_SCHEME.equalsIgnoreCase(split0.file().getScheme())) {
+                HadoopIgfsEndpoint endpoint = new HadoopIgfsEndpoint(split0.file().getAuthority());
+
+                IgfsEx igfs = null;
+
+                if (F.eq(ignite.name(), endpoint.grid()))
+                    igfs = (IgfsEx)((IgniteEx)ignite).igfsx(endpoint.igfs());
+
+                if (igfs != null && !igfs.isProxy(split0.file())) {
+                    IgfsPath path = new IgfsPath(split0.file());
+
+                    if (igfs.exists(path)) {
+                        Collection<IgfsBlockLocation> blocks;
+
+                        try {
+                            blocks = igfs.affinity(path, split0.start(), split0.length());
+                        }
+                        catch (IgniteException e) {
+                            throw new IgniteCheckedException(e);
+                        }
+
+                        assert blocks != null;
+
+                        if (blocks.size() == 1)
+                            // Fast-path, split consists of one IGFS block (as in most cases).
+                            return bestNode(blocks.iterator().next().nodeIds(), topIds, nodeLoads, false);
+                        else {
+                            // Slow-path, file consists of multiple IGFS blocks. First, find the most co-located nodes.
+                            Map<UUID, Long> nodeMap = new HashMap<>();
+
+                            List<UUID> bestNodeIds = null;
+                            long bestLen = -1L;
+
+                            for (IgfsBlockLocation block : blocks) {
+                                for (UUID blockNodeId : block.nodeIds()) {
+                                    if (topIds.contains(blockNodeId)) {
+                                        Long oldLen = nodeMap.get(blockNodeId);
+                                        long newLen = oldLen == null ? block.length() : oldLen + block.length();
+
+                                        nodeMap.put(blockNodeId, newLen);
+
+                                        if (bestNodeIds == null || bestLen < newLen) {
+                                            bestNodeIds = new ArrayList<>(1);
+
+                                            bestNodeIds.add(blockNodeId);
+
+                                            bestLen = newLen;
+                                        }
+                                        else if (bestLen == newLen) {
+                                            assert !F.isEmpty(bestNodeIds);
+
+                                            bestNodeIds.add(blockNodeId);
+                                        }
+                                    }
+                                }
+                            }
+
+                            if (bestNodeIds != null) {
+                                return bestNodeIds.size() == 1 ? bestNodeIds.get(0) :
+                                    bestNode(bestNodeIds, topIds, nodeLoads, true);
+                            }
+                        }
+                    }
+                }
+            }
+        }
+
+        // Cannot use local IGFS for some reason, try selecting the node by host.
+        Collection<UUID> blockNodes = null;
+
+        for (String host : split.hosts()) {
+            Collection<UUID> hostNodes = nodes.get(host);
+
+            if (!F.isEmpty(hostNodes)) {
+                if (blockNodes == null)
+                    blockNodes = new ArrayList<>(hostNodes);
+                else
+                    blockNodes.addAll(hostNodes);
+            }
+        }
+
+        return bestNode(blockNodes, topIds, nodeLoads, false);
+    }
+
+    /**
+     * Finds the best (the least loaded) node among the candidates.
+     *
+     * @param candidates Candidates.
+     * @param topIds Topology node IDs.
+     * @param nodeLoads Known node loads.
+     * @param skipTopCheck Whether to skip topology check.
+     * @return The best node.
+     */
+    private UUID bestNode(@Nullable Collection<UUID> candidates, Collection<UUID> topIds, Map<UUID, Integer> nodeLoads,
+        boolean skipTopCheck) {
+        UUID bestNode = null;
+        int bestLoad = Integer.MAX_VALUE;
+
+        if (candidates != null) {
+            for (UUID candidate : candidates) {
+                if (skipTopCheck || topIds.contains(candidate)) {
+                    int load = nodeLoads.get(candidate);
+
+                    if (bestNode == null || bestLoad > load) {
+                        bestNode = candidate;
+                        bestLoad = load;
+
+                        if (bestLoad == 0)
+                            break; // Minimum load possible, no need for further iterations.
+                    }
+                }
+            }
+        }
+
+        if (bestNode == null) {
+            // Blocks are located on nodes which are not Hadoop-enabled, assign to the least loaded one.
+            bestLoad = Integer.MAX_VALUE;
+
+            for (UUID nodeId : topIds) {
+                int load = nodeLoads.get(nodeId);
+
+                if (bestNode == null || bestLoad > load) {
+                    bestNode = nodeId;
+                    bestLoad = load;
+
+                    if (bestLoad == 0)
+                        break; // Minimum load possible, no need for further iterations.
+                }
+            }
+        }
+
+        assert bestNode != null;
+
+        return bestNode;
+    }
+
+    /**
+     * Create plan for reducers.
+     *
+     * @param top Topology.
+     * @param mappers Mappers map.
+     * @param reducerCnt Reducers count.
+     * @return Reducers map.
+     */
+    private Map<UUID, int[]> reducers(Collection<ClusterNode> top,
+        Map<UUID, Collection<HadoopInputSplit>> mappers, int reducerCnt) {
+        // Determine initial node weights.
+        int totalWeight = 0;
+
+        List<WeightedNode> nodes = new ArrayList<>(top.size());
+
+        for (ClusterNode node : top) {
+            Collection<HadoopInputSplit> split = mappers.get(node.id());
+
+            int weight = reducerNodeWeight(node, split != null ? split.size() : 0);
+
+            nodes.add(new WeightedNode(node.id(), weight, weight));
+
+            totalWeight += weight;
+        }
+
+        // Adjust weights.
+        int totalAdjustedWeight = 0;
+
+        for (WeightedNode node : nodes) {
+            node.floatWeight = ((float)node.weight * reducerCnt) / totalWeight;
+
+            node.weight = Math.round(node.floatWeight);
+
+            totalAdjustedWeight += node.weight;
+        }
+
+        // Apply redundant/lost reducers.
+        Collections.sort(nodes);
+
+        if (totalAdjustedWeight > reducerCnt) {
+            // Too much reducers set.
+            ListIterator<WeightedNode> iter = nodes.listIterator(nodes.size() - 1);
+
+            while (totalAdjustedWeight != reducerCnt) {
+                if (!iter.hasPrevious())
+                    iter = nodes.listIterator(nodes.size() - 1);
+
+                WeightedNode node = iter.previous();
+
+                if (node.weight > 0) {
+                    node.weight -= 1;
+
+                    totalAdjustedWeight--;
+                }
+            }
+        }
+        else if (totalAdjustedWeight < reducerCnt) {
+            // Not enough reducers set.
+            ListIterator<WeightedNode> iter = nodes.listIterator(0);
+
+            while (totalAdjustedWeight != reducerCnt) {
+                if (!iter.hasNext())
+                    iter = nodes.listIterator(0);
+
+                WeightedNode node = iter.next();
+
+                if (node.floatWeight > 0.0f) {
+                    node.weight += 1;
+
+                    totalAdjustedWeight++;
+                }
+            }
+        }
+
+        int idx = 0;
+
+        Map<UUID, int[]> reducers = new HashMap<>(nodes.size(), 1.0f);
+
+        for (WeightedNode node : nodes) {
+            if (node.weight > 0) {
+                int[] arr = new int[node.weight];
+
+                for (int i = 0; i < arr.length; i++)
+                    arr[i] = idx++;
+
+                reducers.put(node.nodeId, arr);
+            }
+        }
+
+        return reducers;
+    }
+
+    /**
+     * Calculate node weight based on node metrics and data co-location.
+     *
+     * @param node Node.
+     * @param splitCnt Splits mapped to this node.
+     * @return Node weight.
+     */
+    @SuppressWarnings("UnusedParameters")
+    protected int reducerNodeWeight(ClusterNode node, int splitCnt) {
+        return splitCnt;
+    }
+
+    /**
+     * Weighted node.
+     */
+    private static class WeightedNode implements Comparable<WeightedNode> {
+        /** Node ID. */
+        private final UUID nodeId;
+
+        /** Weight. */
+        private int weight;
+
+        /** Floating point weight. */
+        private float floatWeight;
+
+        /**
+         * Constructor.
+         *
+         * @param nodeId Node ID.
+         * @param weight Weight.
+         * @param floatWeight Floating point weight.
+         */
+        private WeightedNode(UUID nodeId, int weight, float floatWeight) {
+            this.nodeId = nodeId;
+            this.weight = weight;
+            this.floatWeight = floatWeight;
+        }
+
+        /** {@inheritDoc} */
+        @Override public boolean equals(Object obj) {
+            return obj != null && obj instanceof WeightedNode && F.eq(nodeId, ((WeightedNode)obj).nodeId);
+        }
+
+        /** {@inheritDoc} */
+        @Override public int hashCode() {
+            return nodeId.hashCode();
+        }
+
+        /** {@inheritDoc} */
+        @Override public int compareTo(@NotNull WeightedNode other) {
+            float res = other.floatWeight - floatWeight;
+
+            return res > 0.0f ? 1 : res < 0.0f ? -1 : nodeId.compareTo(other.nodeId);
+        }
+    }
+}
\ No newline at end of file


Mime
View raw message