<?xml version="1.0"?>
<!--
  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.
-->

<!DOCTYPE document PUBLIC "-//APACHE//DTD Documentation V2.0//EN"
          "http://forrest.apache.org/dtd/document-v20.dtd">


<document>

  <header>
    <title>
      HDFS Users Guide
    </title>
  </header>

  <body>
    <section> <title>Purpose</title>
      <p>
 This document is a starting point for users working with
 Hadoop Distributed File System (HDFS) either as a part of a Hadoop cluster  
 or as a stand-alone general purpose distributed file system.
 While HDFS is designed to "just work" in many environments, a working
 knowledge of HDFS helps greatly with configuration improvements and
 diagnostics on a specific cluster.
      </p>
    </section>

    <section> <title> Overview </title>
      <p>
 HDFS is the primary distributed storage used by Hadoop applications. A
 HDFS cluster primarily consists of a NameNode that manages the
 file system metadata and DataNodes that store the actual data. The
 <a href="hdfs_design.html">HDFS Architecture Guide</a> describes HDFS in detail. This user guide primarily deals with 
 the interaction of users and administrators with HDFS clusters. 
 The <a href="images/hdfsarchitecture.gif">HDFS architecture diagram</a> depicts 
 basic interactions among NameNode, the DataNodes, and the clients. 
 Clients contact NameNode for file metadata or file modifications and perform 
 actual file I/O directly with the DataNodes.
      </p>
      <p>
 The following are some of the salient features that could be of
 interest to many users.
      </p> 
    <ul>
    <li>
    	Hadoop, including HDFS, is well suited for distributed storage
    	and distributed processing using commodity hardware. It is fault
    	tolerant, scalable, and extremely simple to expand. MapReduce, 
    	well known for its simplicity and applicability for large set of
    	distributed applications, is an integral part of Hadoop.
    </li>
    <li>
    	HDFS is highly configurable with a default configuration well
    	suited for many installations. Most of the time, configuration
    	needs to be tuned only for very large clusters.
    </li>
    <li>
    	Hadoop is written in Java and is supported on all major platforms.
    </li>
    <li>
    	Hadoop supports shell-like commands to interact with HDFS directly.
    </li>
    <li>
    	The NameNode and Datanodes have built in web servers that makes it
    	easy to check current status of the cluster.
    </li>
    <li>
    	New features and improvements are regularly implemented in HDFS.
    	The following is a subset of useful features in HDFS:
      <ul>
    	<li>
    		File permissions and authentication.
    	</li>
    	<li>
    		<em>Rack awareness</em>: to take a node's physical location into
    		account while scheduling tasks and allocating storage.
    	</li>
    	<li>
    		Safemode: an administrative mode for maintenance.
    	</li>
    	<li>
    		<code>fsck</code>: a utility to diagnose health of the file system, to
    		find missing files or blocks.
    	</li>
    	<li>
    		Rebalancer: tool to balance the cluster when the data is
    		unevenly distributed among DataNodes.
    	</li>
    	<li>
    		Upgrade and rollback: after a software upgrade, 
            it is possible to
    		rollback to HDFS' state before the upgrade in case of unexpected
    		problems.
    	</li>
    	<li>
    		Secondary NameNode (deprecated): performs periodic checkpoints of the 
    		namespace and helps keep the size of file containing log of HDFS 
    		modifications within certain limits at the NameNode.
    		Replaced by Checkpoint node.
    	</li>
    	<li>
    		Checkpoint node: performs periodic checkpoints of the namespace and
    		helps minimize the size of the log stored at the NameNode 
    		containing changes to the HDFS.
    		Replaces the role previously filled by the Secondary NameNode. 
    		NameNode allows multiple Checkpoint nodes simultaneously, 
    		as long as there are no Backup nodes registered with the system.
    	</li>
    	<li>
    		Backup node: An extension to the Checkpoint node.
    		In addition to checkpointing it also receives a stream of edits 
    		from the NameNode and maintains its own in-memory copy of the namespace,
    		which is always in sync with the active NameNode namespace state.
    		Only one Backup node may be registered with the NameNode at once.
    	</li>
      </ul>
    </li>
    </ul>
    
    </section> <section> <title> Prerequisites </title>
    <p>
 	The following documents describe how to install and set up a Hadoop cluster: 
    </p>
 	<ul>
 	<li>
 		<a href="http://hadoop.apache.org/common/docs/current/single_node_setup.html">Single Node Setup</a>
 		for first-time users.
 	</li>
 	<li>
 		<a href="http://hadoop.apache.org/common/docs/current/cluster_setup.html">Cluster Setup</a>
 		for large, distributed clusters.
 	</li>
    </ul>
    <p>
 	The rest of this document assumes the user is able to set up and run a
 	HDFS with at least one DataNode. For the purpose of this document,
 	both the NameNode and DataNode could be running on the same physical
 	machine.	
    </p>
     
    </section> <section> <title> Web Interface </title>
 <p>
 	NameNode and DataNode each run an internal web server in order to
 	display basic information about the current status of the cluster.
 	With the default configuration, the NameNode front page is at
 	<code>http://namenode-name:50070/</code>.
 	It lists the DataNodes in the cluster and basic statistics of the
 	cluster. The web interface can also be used to browse the file
 	system (using "Browse the file system" link on the NameNode front
 	page).
 </p> 
 	
    </section> <section> <title>Shell Commands</title>
 	<p>
      Hadoop includes various shell-like commands that directly
      interact with HDFS and other file systems that Hadoop supports.
      The command
      <code>bin/hdfs dfs -help</code>
      lists the commands supported by Hadoop
      shell. Furthermore, the command
      <code>bin/hdfs dfs -help command-name</code>
      displays more detailed help for a command. These commands support
      most of the normal files system operations like copying files,
      changing file permissions, etc. It also supports a few HDFS
      specific operations like changing replication of files. 
      For more information see <a href="http://hadoop.apache.org/common/docs/current/file_system_shell.html">File System Shell Guide</a>.
     </p>

   <section> <title> DFSAdmin Command </title>
   <p>
   	The <code>bin/hadoop dfsadmin</code>
   	command supports a few HDFS administration related operations.
   	The <code>bin/hadoop dfsadmin -help</code> command
   	lists all the commands currently supported. For e.g.:
   </p>  
   	<ul>
   	<li>
   	    <code>-report</code>
   	    : reports basic statistics of HDFS. Some of this information is
   	    also available on the NameNode front page.
   	</li>
   	<li>
   		<code>-safemode</code>
   		: though usually not required, an administrator can manually enter
   		or leave Safemode.
   	</li>
   	<li>
   		<code>-finalizeUpgrade</code>
   		: removes previous backup of the cluster made during last upgrade.
   	</li>
    <li>
      <code>-refreshNodes</code>
      : Updates the set of hosts allowed to connect to namenode.
      Re-reads the config file to update values defined by dfs.hosts and 
      dfs.host.exclude and reads the entires (hostnames) in those files. 
      Each entry not defined in dfs.hosts but in dfs.hosts.exclude 
      is decommissioned. Each entry defined in dfs.hosts and also in 
      dfs.host.exclude is stopped from decommissioning if it has aleady 
      been marked for decommission. Entires not present in both the lists 
      are decommissioned. 
    </li>
    <li>
      <code>-printTopology</code>
      : Print the topology of the cluster. Display a tree of racks and
      datanodes attached to the tracks as viewed by the NameNode.
    </li>
   	</ul>
   	<p>
   	  For command usage, see  
   	  <a href="http://hadoop.apache.org/common/docs/current/commands_manual.html#dfsadmin">dfsadmin</a>.
   	</p>  
   </section>
   
   </section> 
	<section> <title>Secondary NameNode</title>
   <note>
   The Secondary NameNode has been deprecated. 
   Instead, consider using the 
   <a href="hdfs_user_guide.html#Checkpoint+node">Checkpoint Node</a> or 
   <a href="hdfs_user_guide.html#Backup+node">Backup Node</a>.
   </note>
   <p>	
     The NameNode stores modifications to the file system as a log
     appended to a native file system file, <code>edits</code>. 
   	When a NameNode starts up, it reads HDFS state from an image
   	file, <code>fsimage</code>, and then applies edits from the
    edits log file. It then writes new HDFS state to the <code>fsimage</code>
    and starts normal
   	operation with an empty edits file. Since NameNode merges
   	<code>fsimage</code> and <code>edits</code> files only during start up, 
    the edits log file could get very large over time on a busy cluster. 
    Another side effect of a larger edits file is that next 
    restart of NameNode takes longer.
   </p>
   <p>
     The secondary NameNode merges the fsimage and the edits log files periodically
     and keeps edits log size within a limit. It is usually run on a
     different machine than the primary NameNode since its memory requirements
     are on the same order as the primary NameNode. The secondary
     NameNode is started by <code>bin/start-dfs.sh</code> on the nodes 
     specified in <code>conf/masters</code> file.
   </p>
   <p>
     The start of the checkpoint process on the secondary NameNode is 
     controlled by two configuration parameters.
   </p>
   <ul>
      <li>
        <code>fs.checkpoint.period</code>, set to 1 hour by default, specifies
        the maximum delay between two consecutive checkpoints, and 
      </li>
      <li>
        <code>fs.checkpoint.size</code>, set to 64MB by default, defines the
        size of the edits log file that forces an urgent checkpoint even if 
        the maximum checkpoint delay is not reached.
      </li>
   </ul>
   <p>
     The secondary NameNode stores the latest checkpoint in a  
     directory which is structured the same way as the primary NameNode's
     directory. So that the check pointed image is always ready to be
     read by the primary NameNode if necessary.
   </p>
   <p>
     For command usage, see  
     <a href="http://hadoop.apache.org/common/docs/current/commands_manual.html#secondarynamenode">secondarynamenode</a>.
   </p>
   
   </section><section> <title> Checkpoint Node </title>
   <p>NameNode persists its namespace using two files: <code>fsimage</code>,
      which is the latest checkpoint of the namespace and <code>edits</code>,
      a journal (log) of changes to the namespace since the checkpoint.
      When a NameNode starts up, it merges the <code>fsimage</code> and
      <code>edits</code> journal to provide an up-to-date view of the
      file system metadata.
      The NameNode then overwrites <code>fsimage</code> with the new HDFS state 
      and begins a new <code>edits</code> journal. 
   </p>
   <p>
     The Checkpoint node periodically creates checkpoints of the namespace. 
     It downloads <code>fsimage</code> and <code>edits</code> from the active 
     NameNode, merges them locally, and uploads the new image back to the 
     active NameNode.
     The Checkpoint node usually runs on a different machine than the NameNode
     since its memory requirements are on the same order as the NameNode.
     The Checkpoint node is started by 
     <code>bin/hdfs namenode -checkpoint</code> on the node 
     specified in the configuration file.
   </p>
   <p>The location of the Checkpoint (or Backup) node and its accompanying 
      web interface are configured via the <code>dfs.backup.address</code> 
      and <code>dfs.backup.http.address</code> configuration variables.
	 </p>
   <p>
     The start of the checkpoint process on the Checkpoint node is 
     controlled by two configuration parameters.
   </p>
   <ul>
      <li>
        <code>fs.checkpoint.period</code>, set to 1 hour by default, specifies
        the maximum delay between two consecutive checkpoints 
      </li>
      <li>
        <code>fs.checkpoint.size</code>, set to 64MB by default, defines the
        size of the edits log file that forces an urgent checkpoint even if 
        the maximum checkpoint delay is not reached.
      </li>
   </ul>
   <p>
     The Checkpoint node stores the latest checkpoint in a  
     directory that is structured the same as the NameNode's
     directory. This allows the checkpointed image to be always available for
     reading by the NameNode if necessary.
     See <a href="hdfs_user_guide.html#Import+checkpoint">Import checkpoint</a>.
   </p>
   <p>Multiple checkpoint nodes may be specified in the cluster configuration file.</p>
   <p>
     For command usage, see  
     <a href="http://hadoop.apache.org/common/docs/current/commands_manual.html#namenode">namenode</a>.
   </p>
   </section>

   <section> <title> Backup Node </title>
   <p>	
    The Backup node provides the same checkpointing functionality as the 
    Checkpoint node, as well as maintaining an in-memory, up-to-date copy of the
    file system namespace that is always synchronized with the active NameNode state.
    Along with accepting a journal stream of file system edits from 
    the NameNode and persisting this to disk, the Backup node also applies 
    those edits into its own copy of the namespace in memory, thus creating 
    a backup of the namespace.
   </p>
   <p>
    The Backup node does not need to download 
    <code>fsimage</code> and <code>edits</code> files from the active NameNode
    in order to create a checkpoint, as would be required with a 
    Checkpoint node or Secondary NameNode, since it already has an up-to-date 
    state of the namespace state in memory.
    The Backup node checkpoint process is more efficient as it only needs to 
    save the namespace into the local <code>fsimage</code> file and reset
    <code>edits</code>.
   </p> 
   <p>
    As the Backup node maintains a copy of the
    namespace in memory, its RAM requirements are the same as the NameNode.
   </p> 
   <p>
    The NameNode supports one Backup node at a time. No Checkpoint nodes may be
    registered if a Backup node is in use. Using multiple Backup nodes 
    concurrently will be supported in the future.
   </p> 
   <p>
    The Backup node is configured in the same manner as the Checkpoint node.
    It is started with <code>bin/hdfs namenode -checkpoint</code>.
   </p>
   <p>The location of the Backup (or Checkpoint) node and its accompanying 
      web interface are configured via the <code>dfs.backup.address</code> 
      and <code>dfs.backup.http.address</code> configuration variables.
	 </p>
   <p>
    Use of a Backup node provides the option of running the NameNode with no 
    persistent storage, delegating all responsibility for persisting the state
    of the namespace to the Backup node. 
    To do this, start the NameNode with the 
    <code>-importCheckpoint</code> option, along with specifying no persistent
    storage directories of type edits <code>dfs.name.edits.dir</code> 
    for the NameNode configuration.
   </p> 
   <p>
    For a complete discussion of the motivation behind the creation of the 
    Backup node and Checkpoint node, see 
    <a href="https://issues.apache.org/jira/browse/HADOOP-4539">HADOOP-4539</a>.
    For command usage, see  
     <a href="http://hadoop.apache.org/common/docs/current/commands_manual.html#namenode">namenode</a>.
   </p>
   </section>

   <section> <title> Import Checkpoint </title>
   <p>
     The latest checkpoint can be imported to the NameNode if
     all other copies of the image and the edits files are lost.
     In order to do that one should:
   </p>
   <ul>
      <li>
        Create an empty directory specified in the 
        <code>dfs.name.dir</code> configuration variable;
      </li>
      <li>
        Specify the location of the checkpoint directory in the 
        configuration variable <code>fs.checkpoint.dir</code>;
      </li>
      <li>
        and start the NameNode with <code>-importCheckpoint</code> option.
      </li>
   </ul>
   <p>
     The NameNode will upload the checkpoint from the 
     <code>fs.checkpoint.dir</code> directory and then save it to the NameNode
     directory(s) set in <code>dfs.name.dir</code>.
     The NameNode will fail if a legal image is contained in 
     <code>dfs.name.dir</code>.
     The NameNode verifies that the image in <code>fs.checkpoint.dir</code> is
     consistent, but does not modify it in any way.
   </p>
   <p>
     For command usage, see  
      <a href="http://hadoop.apache.org/common/docs/current/commands_manual.html#namenode">namenode</a>.
   </p>
   </section>

   <section> <title> Rebalancer </title>
    <p>
      HDFS data might not always be be placed uniformly across the
      DataNode. One common reason is addition of new DataNodes to an
      existing cluster. While placing new blocks (data for a file is
      stored as a series of blocks), NameNode considers various
      parameters before choosing the DataNodes to receive these blocks.
      Some of the considerations are: 
    </p>
      <ul>
      <li>
        Policy to keep one of the replicas of a block on the same node
        as the node that is writing the block.
      </li>
      <li>
        Need to spread different replicas of a block across the racks so
        that cluster can survive loss of whole rack.
      </li>
      <li>
        One of the replicas is usually placed on the same rack as the
        node writing to the file so that cross-rack network I/O is
        reduced.
      </li>
      <li>
        Spread HDFS data uniformly across the DataNodes in the cluster.
      </li>
      </ul>
    <p>
      Due to multiple competing considerations, data might not be
      uniformly placed across the DataNodes.
      HDFS provides a tool for administrators that analyzes block
      placement and rebalanaces data across the DataNode. A brief
      administrator's guide for rebalancer as a
      <a href="http://issues.apache.org/jira/secure/attachment/12368261/RebalanceDesign6.pdf">PDF</a>
      is attached to
      <a href="http://issues.apache.org/jira/browse/HADOOP-1652">HADOOP-1652</a>.
    </p>
    <p>
     For command usage, see  
     <a href="http://hadoop.apache.org/common/docs/current/commands_manual.html#balancer">balancer</a>.
   </p>
    
   </section> <section> <title> Rack Awareness </title>
    <p>
      Typically large Hadoop clusters are arranged in racks and
      network traffic between different nodes with in the same rack is
      much more desirable than network traffic across the racks. In
      addition NameNode tries to place replicas of block on
      multiple racks for improved fault tolerance. Hadoop lets the
      cluster administrators decide which rack a node belongs to
      through configuration variable <code>dfs.network.script</code>. When this
      script is configured, each node runs the script to determine its
      rack id. A default installation assumes all the nodes belong to
      the same rack. This feature and configuration is further described
      in <a href="http://issues.apache.org/jira/secure/attachment/12345251/Rack_aware_HDFS_proposal.pdf">PDF</a>
      attached to 
      <a href="http://issues.apache.org/jira/browse/HADOOP-692">HADOOP-692</a>.
    </p>

   </section> <section> <title> Safemode </title>
    <p>
      During start up the NameNode loads the file system state from the
      fsimage  and the edits log file. It then waits for DataNodes
      to report their blocks so that it does not prematurely start
      replicating the blocks though enough replicas already exist in the
      cluster. During this time NameNode stays in Safemode. 
      Safemode
      for the NameNode is essentially a read-only mode for the HDFS cluster,
      where it does not allow any modifications to file system or blocks.
      Normally the NameNode leaves Safemode automatically after the DataNodes
      have reported that most file system blocks are available. 
      If required, HDFS could be placed in Safemode explicitly
      using <code>'bin/hadoop dfsadmin -safemode'</code> command. NameNode front
      page shows whether Safemode is on or off. A more detailed
      description and configuration is maintained as JavaDoc for
      <a href="http://hadoop.apache.org/core/docs/current/api/org/apache/hadoop/dfs/NameNode.html#setSafeMode(org.apache.hadoop.dfs.FSConstants.SafeModeAction)"><code>setSafeMode()</code></a>.
    </p>
    
   </section> <section> <title> fsck </title>
     <p>    
      HDFS supports the <code>fsck</code> command to check for various 
      inconsistencies.
      It it is designed for reporting problems with various
      files, for example, missing blocks for a file or under-replicated
      blocks. Unlike a traditional <code>fsck</code> utility for native file systems,
      this command does not correct the errors it detects. Normally NameNode
      automatically corrects most of the recoverable failures. By default
      <code>fsck</code> ignores open files but provides an option to select all files during reporting.
      The HDFS <code>fsck</code> command is not a
      Hadoop shell command. It can be run as '<code>bin/hadoop fsck</code>'.
      For command usage, see  
      <a href="http://hadoop.apache.org/common/docs/current/commands_manual.html#fsck">fsck</a>.
      <code>fsck</code> can be run on the whole file system or on a subset of files.
     </p>
     
   </section> <section> <title> Upgrade and Rollback </title>
     <p>
      When Hadoop is upgraded on an existing cluster, as with any
      software upgrade, it is possible there are new bugs or
      incompatible changes that affect existing applications and were
      not discovered earlier. In any non-trivial HDFS installation, it
      is not an option to loose any data, let alone to restart HDFS from
      scratch. HDFS allows administrators to go back to earlier version
      of Hadoop and rollback the cluster to the state it was in 
      before
      the upgrade. HDFS upgrade is described in more detail in 
      <a href="http://wiki.apache.org/hadoop/Hadoop_Upgrade">Hadoop Upgrade</a> Wiki page.
      HDFS can have one such backup at a time. Before upgrading,
      administrators need to remove existing backup using <code>bin/hadoop
      dfsadmin -finalizeUpgrade</code> command. The following
      briefly describes the typical upgrade procedure: 
     </p>
      <ul>
      <li>
        Before upgrading Hadoop software,
        <em>finalize</em> if there an existing backup.
        <code>dfsadmin -upgradeProgress status</code>
        can tell if the cluster needs to be <em>finalized</em>.
      </li>
      <li>Stop the cluster and distribute new version of Hadoop.</li>
      <li>
        Run the new version with <code>-upgrade</code> option 
        (<code>bin/start-dfs.sh -upgrade</code>).
      </li>
      <li>
        Most of the time, cluster works just fine. Once the new HDFS is
        considered working well (may be after a few days of operation),
        finalize the upgrade. Note that until the cluster is finalized,
        deleting the files that existed before the upgrade does not free
        up real disk space on the DataNodes.
      </li>
      <li>
        If there is a need to move back to the old version,
        <ul>
          <li> stop the cluster and distribute earlier version of Hadoop. </li>
          <li> start the cluster with rollback option. 
            (<code>bin/start-dfs.h -rollback</code>). 
          </li>
        </ul>
      </li>
      </ul>
    
   </section> <section> <title> File Permissions and Security </title>
     <p>           
      The file permissions are designed to be similar to file permissions on
      other familiar platforms like Linux. Currently, security is limited
      to simple file permissions. The user that starts NameNode is
      treated as the superuser for HDFS. Future versions of HDFS will
      support network authentication protocols like Kerberos for user
      authentication and encryption of data transfers. The details are discussed in the 
      <a href="hdfs_permissions_guide.html">Permissions Guide</a>.
     </p>
     
   </section> <section> <title> Scalability </title>
     <p>
      Hadoop currently runs on clusters with thousands of nodes. The  
      <a href="http://wiki.apache.org/hadoop/PoweredBy">PoweredBy</a> Wiki page 
      lists some of the organizations that deploy Hadoop on large
      clusters. HDFS has one NameNode for each cluster. Currently
      the total memory available on NameNode is the primary scalability
      limitation. On very large clusters, increasing average size of
      files stored in HDFS helps with increasing cluster size without
      increasing memory requirements on NameNode.
   
      The default configuration may not suite very large clustes. The 
      <a href="http://wiki.apache.org/hadoop/FAQ">FAQ</a> Wiki page lists
      suggested configuration improvements for large Hadoop clusters.
     </p>
     
   </section> <section> <title> Related Documentation </title>          
      <p>
      This user guide is a good starting point for
      working with HDFS. While the user guide continues to improve,
      there is a large wealth of documentation about Hadoop and HDFS.
      The following list is a starting point for further exploration:
      </p>
      <ul>
      <li>
        <a href="http://hadoop.apache.org/">Hadoop Site</a>: The home page for the Apache Hadoop site.
      </li>
      <li>
        <a href="http://wiki.apache.org/hadoop/FrontPage">Hadoop Wiki</a>:
        The home page (FrontPage) for the Hadoop Wiki. Unlike the released documentation, 
        which is part of Hadoop source tree, Hadoop Wiki is
        regularly edited by Hadoop Community.
      </li>
      <li> <a href="http://wiki.apache.org/hadoop/FAQ">FAQ</a>: 
      The FAQ Wiki page.
      </li>
      <li>
        Hadoop <a href="http://hadoop.apache.org/core/docs/current/api/">
          JavaDoc API</a>.
      </li>
      <li>
        Hadoop User Mailing List : 
        <a href="mailto:core-user@hadoop.apache.org">core-user[at]hadoop.apache.org</a>.
      </li>
      <li>
         Explore <code>src/hdfs/hdfs-default.xml</code>. 
         It includes brief 
         description of most of the configuration variables available.
      </li>
      <li>
        <a href="http://hadoop.apache.org/common/docs/current/commands_manual.html">Hadoop Commands Guide</a>: Hadoop commands usage.
      </li>
      </ul>
     </section>
     
  </body>
</document>