hadoop 我的配置

slave 

192.168.1.101

192.168.1.102

192.168.1.103

192.168.1.104

 

master

192.168.1.3

 

 

core-site.xml

<?xml version="1.0"?>

<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>

 

<!-- Do not modify this file directly.  Instead, copy entries that you -->

<!-- wish to modify from this file into core-site.xml and change them -->

<!-- there.  If core-site.xml does not already exist, create it.      -->

 

<configuration>

 

<!--- global properties -->

 

<property>

  <name>hadoop.tmp.dir</name>

  <value>/tmp/hadoop-${user.name}</value>

  <description>A base for other temporary directories.</description>

</property>

 

<property>

  <name>hadoop.native.lib</name>

  <value>true</value>

  <description>Should native hadoop libraries, if present, be used.</description>

</property>

 

<property>

  <name>hadoop.http.filter.initializers</name>

  <value></value>

  <description>A comma separated list of class names. Each class in the list 

  must extend org.apache.hadoop.http.FilterInitializer. The corresponding 

  Filter will be initialized. Then, the Filter will be applied to all user 

  facing jsp and servlet web pages.  The ordering of the list defines the 

  ordering of the filters.</description>

</property>

 

<property>

  <name>hadoop.security.authorization</name>

  <value>false</value>

  <description>Is service-level authorization enabled?</description>

</property>

 

<!--- logging properties -->

 

<property>

  <name>hadoop.logfile.size</name>

  <value>10000000</value>

  <description>The max size of each log file</description>

</property>

 

<property>

  <name>hadoop.logfile.count</name>

  <value>10</value>

  <description>The max number of log files</description>

</property>

 

<!-- i/o properties -->

<property>

  <name>io.file.buffer.size</name>

  <value>4096</value>

  <description>The size of buffer for use in sequence files.

  The size of this buffer should probably be a multiple of hardware

  page size (4096 on Intel x86), and it determines how much data is

  buffered during read and write operations.</description>

</property>

 

<property>

  <name>io.bytes.per.checksum</name>

  <value>512</value>

  <description>The number of bytes per checksum.  Must not be larger than

  io.file.buffer.size.</description>

</property>

 

<property>

  <name>io.skip.checksum.errors</name>

  <value>false</value>

  <description>If true, when a checksum error is encountered while

  reading a sequence file, entries are skipped, instead of throwing an

  exception.</description>

</property>

 

<property>

  <name>io.compression.codecs</name>

  <value>org.apache.hadoop.io.compress.DefaultCodec,org.apache.hadoop.io.compress.GzipCodec,org.apache.hadoop.io.compress.BZip2Codec</value>

  <description>A list of the compression codec classes that can be used 

               for compression/decompression.</description>

</property>

 

<property>

  <name>io.serializations</name>

  <value>org.apache.hadoop.io.serializer.WritableSerialization</value>

  <description>A list of serialization classes that can be used for

  obtaining serializers and deserializers.</description>

</property>

 

<!-- file system properties -->

 

<property>

  <name>fs.default.name</name>

  <value>hdfs://192.168.1.3:9000</value>

  <description>The name of the default file system.  A URI whose

  scheme and authority determine the FileSystem implementation.  The

  uri's scheme determines the config property (fs.SCHEME.impl) naming

  the FileSystem implementation class.  The uri's authority is used to

  determine the host, port, etc. for a filesystem.</description>

</property>

 

<property>

  <name>fs.trash.interval</name>

  <value>0</value>

  <description>Number of minutes between trash checkpoints.

  If zero, the trash feature is disabled.

  </description>

</property>

 

<property>

  <name>fs.file.impl</name>

  <value>org.apache.hadoop.fs.LocalFileSystem</value>

  <description>The FileSystem for file: uris.</description>

</property>

 

<property>

  <name>fs.hdfs.impl</name>

  <value>org.apache.hadoop.hdfs.DistributedFileSystem</value>

  <description>The FileSystem for hdfs: uris.</description>

</property>

 

<property>

  <name>fs.s3.impl</name>

  <value>org.apache.hadoop.fs.s3.S3FileSystem</value>

  <description>The FileSystem for s3: uris.</description>

</property>

 

<property>

  <name>fs.s3n.impl</name>

  <value>org.apache.hadoop.fs.s3native.NativeS3FileSystem</value>

  <description>The FileSystem for s3n: (Native S3) uris.</description>

</property>

 

<property>

  <name>fs.kfs.impl</name>

  <value>org.apache.hadoop.fs.kfs.KosmosFileSystem</value>

  <description>The FileSystem for kfs: uris.</description>

</property>

 

<property>

  <name>fs.hftp.impl</name>

  <value>org.apache.hadoop.hdfs.HftpFileSystem</value>

</property>

 

<property>

  <name>fs.hsftp.impl</name>

  <value>org.apache.hadoop.hdfs.HsftpFileSystem</value>

</property>

 

<property>

  <name>fs.ftp.impl</name>

  <value>org.apache.hadoop.fs.ftp.FTPFileSystem</value>

  <description>The FileSystem for ftp: uris.</description>

</property>

 

<property>

  <name>fs.ramfs.impl</name>

  <value>org.apache.hadoop.fs.InMemoryFileSystem</value>

  <description>The FileSystem for ramfs: uris.</description>

</property>

 

<property>

  <name>fs.har.impl</name>

  <value>org.apache.hadoop.fs.HarFileSystem</value>

  <description>The filesystem for Hadoop archives. </description>

</property>

 

<property>

  <name>fs.har.impl.disable.cache</name>

  <value>true</value>

  <description>Don't cache 'har' filesystem instances.</description>

</property>

 

<property>

  <name>fs.checkpoint.dir</name>

  <value>${hadoop.tmp.dir}/dfs/namesecondary</value>

  <description>Determines where on the local filesystem the DFS secondary

      name node should store the temporary images to merge.

      If this is a comma-delimited list of directories then the image is

      replicated in all of the directories for redundancy.

  </description>

</property>

 

<property>

  <name>fs.checkpoint.edits.dir</name>

  <value>${fs.checkpoint.dir}</value>

  <description>Determines where on the local filesystem the DFS secondary

      name node should store the temporary edits to merge.

      If this is a comma-delimited list of directoires then teh edits is

      replicated in all of the directoires for redundancy.

      Default value is same as fs.checkpoint.dir

  </description>

</property>

 

<property>

  <name>fs.checkpoint.period</name>

  <value>3600</value>

  <description>The number of seconds between two periodic checkpoints.

  </description>

</property>

 

<property>

  <name>fs.checkpoint.size</name>

  <value>67108864</value>

  <description>The size of the current edit log (in bytes) that triggers

       a periodic checkpoint even if the fs.checkpoint.period hasn't expired.

  </description>

</property>

 

 

 

<property>

  <name>fs.s3.block.size</name>

  <value>67108864</value>

  <description>Block size to use when writing files to S3.</description>

</property>

 

<property>

  <name>fs.s3.buffer.dir</name>

  <value>${hadoop.tmp.dir}/s3</value>

  <description>Determines where on the local filesystem the S3 filesystem

  should store files before sending them to S3

  (or after retrieving them from S3).

  </description>

</property>

 

<property>

  <name>fs.s3.maxRetries</name>

  <value>4</value>

  <description>The maximum number of retries for reading or writing files to S3, 

  before we signal failure to the application.

  </description>

</property>

 

<property>

  <name>fs.s3.sleepTimeSeconds</name>

  <value>10</value>

  <description>The number of seconds to sleep between each S3 retry.

  </description>

</property>

 

 

<property>

  <name>local.cache.size</name>

  <value>10737418240</value>

  <description>The limit on the size of cache you want to keep, set by default

  to 10GB. This will act as a soft limit on the cache directory for out of band data.

  </description>

</property>

 

<property>

  <name>io.seqfile.compress.blocksize</name>

  <value>1000000</value>

  <description>The minimum block size for compression in block compressed 

          SequenceFiles.

  </description>

</property>

 

<property>

  <name>io.seqfile.lazydecompress</name>

  <value>true</value>

  <description>Should values of block-compressed SequenceFiles be decompressed

          only when necessary.

  </description>

</property>

 

<property>

  <name>io.seqfile.sorter.recordlimit</name>

  <value>1000000</value>

  <description>The limit on number of records to be kept in memory in a spill 

          in SequenceFiles.Sorter

  </description>

</property>

 

 <property>

  <name>io.mapfile.bloom.size</name>

  <value>1048576</value>

  <description>The size of BloomFilter-s used in BloomMapFile. Each time this many

  keys is appended the next BloomFilter will be created (inside a DynamicBloomFilter).

  Larger values minimize the number of filters, which slightly increases the performance,

  but may waste too much space if the total number of keys is usually much smaller

  than this number.

  </description>

</property>

 

<property>

  <name>io.mapfile.bloom.error.rate</name>

  <value>0.005</value>

  <description>The rate of false positives in BloomFilter-s used in BloomMapFile.

  As this value decreases, the size of BloomFilter-s increases exponentially. This

  value is the probability of encountering false positives (default is 0.5%).

  </description>

</property>

 

<property>

  <name>hadoop.util.hash.type</name>

  <value>murmur</value>

  <description>The default implementation of Hash. Currently this can take one of the

  two values: 'murmur' to select MurmurHash and 'jenkins' to select JenkinsHash.

  </description>

</property>

 

 

<!-- ipc properties -->

 

<property>

  <name>ipc.client.idlethreshold</name>

  <value>4000</value>

  <description>Defines the threshold number of connections after which

               connections will be inspected for idleness.

  </description>

</property>

 

<property>

  <name>ipc.client.kill.max</name>

  <value>10</value>

  <description>Defines the maximum number of clients to disconnect in one go.

  </description>

</property>

 

<property>

  <name>ipc.client.connection.maxidletime</name>

  <value>10000</value>

  <description>The maximum time in msec after which a client will bring down the

               connection to the server.

  </description>

</property>

 

<property>

  <name>ipc.client.connect.max.retries</name>

  <value>10</value>

  <description>Indicates the number of retries a client will make to establish

               a server connection.

  </description>

</property>

 

<property>

  <name>ipc.server.listen.queue.size</name>

  <value>128</value>

  <description>Indicates the length of the listen queue for servers accepting

               client connections.

  </description>

</property>

 

<property>

  <name>ipc.server.tcpnodelay</name>

  <value>false</value>

  <description>Turn on/off Nagle's algorithm for the TCP socket connection on 

  the server. Setting to true disables the algorithm and may decrease latency

  with a cost of more/smaller packets. 

  </description>

</property>

 

<property>

  <name>ipc.client.tcpnodelay</name>

  <value>false</value>

  <description>Turn on/off Nagle's algorithm for the TCP socket connection on 

  the client. Setting to true disables the algorithm and may decrease latency

  with a cost of more/smaller packets. 

  </description>

</property>

 

 

<!-- Web Interface Configuration -->

 

<property>

  <name>webinterface.private.actions</name>

  <value>false</value>

  <description> If set to true, the web interfaces of JT and NN may contain 

                actions, such as kill job, delete file, etc., that should 

                not be exposed to public. Enable this option if the interfaces 

                are only reachable by those who have the right authorization.

  </description>

</property>

 

<!-- Proxy Configuration -->

 

<property>

  <name>hadoop.rpc.socket.factory.class.default</name>

  <value>org.apache.hadoop.net.StandardSocketFactory</value>

  <description> Default SocketFactory to use. This parameter is expected to be

    formatted as "package.FactoryClassName".

  </description>

</property>

 

<property>

  <name>hadoop.rpc.socket.factory.class.ClientProtocol</name>

  <value></value>

  <description> SocketFactory to use to connect to a DFS. If null or empty, use

    hadoop.rpc.socket.class.default. This socket factory is also used by

    DFSClient to create sockets to DataNodes.

  </description>

</property>

 

 

 

<property>

  <name>hadoop.socks.server</name>

  <value></value>

  <description> Address (host:port) of the SOCKS server to be used by the

    SocksSocketFactory.

  </description>

</property>

 

<!-- Rack Configuration -->

 

<property>

  <name>topology.node.switch.mapping.impl</name>

  <value>org.apache.hadoop.net.ScriptBasedMapping</value>

  <description> The default implementation of the DNSToSwitchMapping. It

    invokes a script specified in topology.script.file.name to resolve

    node names. If the value for topology.script.file.name is not set, the

    default value of DEFAULT_RACK is returned for all node names.

  </description>

</property>

 

<property>

  <name>topology.script.file.name</name>

  <value></value>

  <description> The script name that should be invoked to resolve DNS names to

    NetworkTopology names. Example: the script would take host.foo.bar as an

    argument, and return /rack1 as the output.

  </description>

</property>

 

<property>

  <name>topology.script.number.args</name>

  <value>100</value>

  <description> The max number of args that the script configured with 

    topology.script.file.name should be run with. Each arg is an

    IP address.

  </description>

</property>

 

 

 

</configuration>

 

 

 

 

 

 

 

 

 

 

hdfs-site.xml

<?xml version="1.0"?>

<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>

 

<!-- Do not modify this file directly.  Instead, copy entries that you -->

<!-- wish to modify from this file into hdfs-site.xml and change them -->

<!-- there.  If hdfs-site.xml does not already exist, create it.      -->

 

<configuration>

 

<property>

  <name>dfs.namenode.logging.level</name>

  <value>info</value>

  <description>The logging level for dfs namenode. Other values are "dir"(trac

e namespace mutations), "block"(trace block under/over replications and block

creations/deletions), or "all".</description>

</property>

 

<property>

  <name>dfs.secondary.http.address</name>

  <value>0.0.0.0:50090</value>

  <description>

    The secondary namenode http server address and port.

    If the port is 0 then the server will start on a free port.

  </description>

</property>

 

<property>

  <name>dfs.datanode.address</name>

  <value>0.0.0.0:50010</value>

  <description>

    The address where the datanode server will listen to.

    If the port is 0 then the server will start on a free port.

  </description>

</property>

 

<property>

  <name>dfs.datanode.http.address</name>

  <value>0.0.0.0:50075</value>

  <description>

    The datanode http server address and port.

    If the port is 0 then the server will start on a free port.

  </description>

</property>

 

<property>

  <name>dfs.datanode.ipc.address</name>

  <value>0.0.0.0:50020</value>

  <description>

    The datanode ipc server address and port.

    If the port is 0 then the server will start on a free port.

  </description>

</property>

 

<property>

  <name>dfs.datanode.handler.count</name>

  <value>3</value>

  <description>The number of server threads for the datanode.</description>

</property>

 

<property>

  <name>dfs.http.address</name>

  <value>0.0.0.0:50070</value>

  <description>

    The address and the base port where the dfs namenode web ui will listen on.

    If the port is 0 then the server will start on a free port.

  </description>

</property>

 

<property>

  <name>dfs.https.enable</name>

  <value>false</value>

  <description>Decide if HTTPS(SSL) is supported on HDFS

  </description>

</property>

 

<property>

  <name>dfs.https.need.client.auth</name>

  <value>false</value>

  <description>Whether SSL client certificate authentication is required

  </description>

</property>

 

<property>

  <name>dfs.https.server.keystore.resource</name>

  <value>ssl-server.xml</value>

  <description>Resource file from which ssl server keystore

  information will be extracted

  </description>

</property>

 

<property>

  <name>dfs.https.client.keystore.resource</name>

  <value>ssl-client.xml</value>

  <description>Resource file from which ssl client keystore

  information will be extracted

  </description>

</property>

 

<property>

  <name>dfs.datanode.https.address</name>

  <value>0.0.0.0:50475</value>

</property>

 

<property>

  <name>dfs.https.address</name>

  <value>0.0.0.0:50470</value>

</property>

 

 <property>

  <name>dfs.datanode.dns.interface</name>

  <value>default</value>

  <description>The name of the Network Interface from which a data node should 

  report its IP address.

  </description>

 </property>

 

<property>

  <name>dfs.datanode.dns.nameserver</name>

  <value>default</value>

  <description>The host name or IP address of the name server (DNS)

  which a DataNode should use to determine the host name used by the

  NameNode for communication and display purposes.

  </description>

 </property>

 

 

 

<property>

  <name>dfs.replication.considerLoad</name>

  <value>true</value>

  <description>Decide if chooseTarget considers the target's load or not

  </description>

</property>

<property>

  <name>dfs.default.chunk.view.size</name>

  <value>32768</value>

  <description>The number of bytes to view for a file on the browser.

  </description>

</property>

 

<property>

  <name>dfs.datanode.du.reserved</name>

  <value>0</value>

  <description>Reserved space in bytes per volume. Always leave this much space free for non dfs use.

  </description>

</property>

 

<property>

  <name>dfs.name.dir</name>

  <value>/home/hadoop/data/dfs.name.dir</value>

  <description>Determines where on the local filesystem the DFS name node

      should store the name table(fsimage).  If this is a comma-delimited list

      of directories then the name table is replicated in all of the

      directories, for redundancy. </description>

</property>

 

<property>

  <name>dfs.name.edits.dir</name>

  <value>${dfs.name.dir}</value>

  <description>Determines where on the local filesystem the DFS name node

      should store the transaction (edits) file. If this is a comma-delimited list

      of directories then the transaction file is replicated in all of the 

      directories, for redundancy. Default value is same as dfs.name.dir

  </description>

</property>

<property>

  <name>dfs.web.ugi</name>

  <value>webuser,webgroup</value>

  <description>The user account used by the web interface.

    Syntax: USERNAME,GROUP1,GROUP2, ...

  </description>

</property>

 

<property>

  <name>dfs.permissions</name>

  <value>false</value>

  <description>

    If "true", enable permission checking in HDFS.

    If "false", permission checking is turned off,

    but all other behavior is unchanged.

    Switching from one parameter value to the other does not change the mode,

    owner or group of files or directories.

  </description>

</property>

 

<property>

  <name>dfs.permissions.supergroup</name>

  <value>supergroup</value>

  <description>The name of the group of super-users.</description>

</property>

 

<property>

  <name>dfs.data.dir</name>

  <value>/home/hadoop/data/dfs.data.dir</value>

  <description>Determines where on the local filesystem an DFS data node

  should store its blocks.  If this is a comma-delimited

  list of directories, then data will be stored in all named

  directories, typically on different devices.

  Directories that do not exist are ignored.

  </description>

</property>

 

<property>

  <name>dfs.replication</name>

  <value>3</value>

  <description>Default block replication. 

  The actual number of replications can be specified when the file is created.

  The default is used if replication is not specified in create time.

  </description>

</property>

 

<property>

  <name>dfs.replication.max</name>

  <value>512</value>

  <description>Maximal block replication. 

  </description>

</property>

 

<property>

  <name>dfs.replication.min</name>

  <value>1</value>

  <description>Minimal block replication. 

  </description>

</property>

 

<property>

  <name>dfs.block.size</name>

  <value>67108864</value>

  <description>The default block size for new files.</description>

</property>

 

<property>

  <name>dfs.df.interval</name>

  <value>60000</value>

  <description>Disk usage statistics refresh interval in msec.</description>

</property>

 

<property>

  <name>dfs.client.block.write.retries</name>

  <value>3</value>

  <description>The number of retries for writing blocks to the data nodes, 

  before we signal failure to the application.

  </description>

</property>

 

<property>

  <name>dfs.blockreport.intervalMsec</name>

  <value>3600000</value>

  <description>Determines block reporting interval in milliseconds.</description>

</property>

 

<property>

  <name>dfs.blockreport.initialDelay</name>  <value>0</value>

  <description>Delay for first block report in seconds.</description>

</property>

 

<property>

  <name>dfs.heartbeat.interval</name>

  <value>3</value>

  <description>Determines datanode heartbeat interval in seconds.</description>

</property>

 

<property>

  <name>dfs.namenode.handler.count</name>

  <value>10</value>

  <description>The number of server threads for the namenode.</description>

</property>

 

<property>

  <name>dfs.safemode.threshold.pct</name>

  <value>0.999f</value>

  <description>

    Specifies the percentage of blocks that should satisfy 

    the minimal replication requirement defined by dfs.replication.min.

    Values less than or equal to 0 mean not to start in safe mode.

    Values greater than 1 will make safe mode permanent.

  </description>

</property>

 

<property>

  <name>dfs.safemode.extension</name>

  <value>30000</value>

  <description>

    Determines extension of safe mode in milliseconds 

    after the threshold level is reached.

  </description>

</property>

 

<property>

  <name>dfs.balance.bandwidthPerSec</name>

  <value>1048576</value>

  <description>

        Specifies the maximum amount of bandwidth that each datanode

        can utilize for the balancing purpose in term of

        the number of bytes per second.

  </description>

</property>

 

<property>

  <name>dfs.hosts</name>

  <value></value>

  <description>Names a file that contains a list of hosts that are

  permitted to connect to the namenode. The full pathname of the file

  must be specified.  If the value is empty, all hosts are

  permitted.</description>

</property>

 

<property>

  <name>dfs.hosts.exclude</name>

  <value></value>

  <description>Names a file that contains a list of hosts that are

  not permitted to connect to the namenode.  The full pathname of the

  file must be specified.  If the value is empty, no hosts are

  excluded.</description>

</property> 

 

<property>

  <name>dfs.max.objects</name>

  <value>0</value>

  <description>The maximum number of files, directories and blocks

  dfs supports. A value of zero indicates no limit to the number

  of objects that dfs supports.

  </description>

</property>

 

<property>

  <name>dfs.namenode.decommission.interval</name>

  <value>30</value>

  <description>Namenode periodicity in seconds to check if decommission is 

  complete.</description>

</property>

 

<property>

  <name>dfs.namenode.decommission.nodes.per.interval</name>

  <value>5</value>

  <description>The number of nodes namenode checks if decommission is complete

  in each dfs.namenode.decommission.interval.</description>

</property>

 

<property>

  <name>dfs.replication.interval</name>

  <value>3</value>

  <description>The periodicity in seconds with which the namenode computes 

  repliaction work for datanodes. </description>

</property>

 

<property>

  <name>dfs.access.time.precision</name>

  <value>3600000</value>

  <description>The access time for HDFS file is precise upto this value. 

               The default value is 1 hour. Setting a value of 0 disables

               access times for HDFS.

  </description>

</property>

 

<property>

  <name>dfs.support.append</name>

  <value>false</value>

  <description>Does HDFS allow appends to files?

               This is currently set to false because there are bugs in the

               "append code" and is not supported in any prodction cluster.

  </description>

</property>

 

</configuration>

 

 

 

 

 

 

mapred-site.xml

<?xml version="1.0"?>

<?xml-stylesheet type="text/xsl" href="configuration.xsl"?>

 

<!-- Do not modify this file directly.  Instead, copy entries that you -->

<!-- wish to modify from this file into mapred-site.xml and change them -->

<!-- there.  If mapred-site.xml does not already exist, create it.      -->

 

<configuration>

 

<property>

  <name>hadoop.job.history.location</name>

  <value></value>

  <description> If job tracker is static the history files are stored 

  in this single well known place. If No value is set here, by default,

  it is in the local file system at ${hadoop.log.dir}/history.

  </description>

</property>

 

<property>

  <name>hadoop.job.history.user.location</name>

  <value></value>

  <description> User can specify a location to store the history files of 

  a particular job. If nothing is specified, the logs are stored in 

  output directory. The files are stored in "_logs/history/" in the directory.

  User can stop logging by giving the value "none". 

  </description>

</property>

 

<!-- i/o properties -->

 

<property>

  <name>io.sort.factor</name>

  <value>10</value>

  <description>The number of streams to merge at once while sorting

  files.  This determines the number of open file handles.</description>

</property>

 

<property>

  <name>io.sort.mb</name>

  <value>100</value>

  <description>The total amount of buffer memory to use while sorting 

  files, in megabytes.  By default, gives each merge stream 1MB, which

  should minimize seeks.</description>

</property>

 

<property>

  <name>io.sort.record.percent</name>

  <value>0.05</value>

  <description>The percentage of io.sort.mb dedicated to tracking record

  boundaries. Let this value be r, io.sort.mb be x. The maximum number

  of records collected before the collection thread must block is equal

  to (r * x) / 4</description>

</property>

 

<property>

  <name>io.sort.spill.percent</name>

  <value>0.80</value>

  <description>The soft limit in either the buffer or record collection

  buffers. Once reached, a thread will begin to spill the contents to disk

  in the background. Note that this does not imply any chunking of data to

  the spill. A value less than 0.5 is not recommended.</description>

</property>

 

<property>

  <name>io.map.index.skip</name>

  <value>0</value>

  <description>Number of index entries to skip between each entry.

  Zero by default. Setting this to values larger than zero can

  facilitate opening large map files using less memory.</description>

</property>

 

<property>

  <name>mapred.job.tracker</name>

  <value>192.168.1.3:9001</value>

  <description>The host and port that the MapReduce job tracker runs

  at.  If "local", then jobs are run in-process as a single map

  and reduce task.

  </description>

</property>

 

<property>

  <name>mapred.job.tracker.http.address</name>

  <value>0.0.0.0:50030</value>

  <description>

    The job tracker http server address and port the server will listen on.

    If the port is 0 then the server will start on a free port.

  </description>

</property>

 

<property>

  <name>mapred.job.tracker.handler.count</name>

  <value>10</value>

  <description>

    The number of server threads for the JobTracker. This should be roughly

    4% of the number of tasktracker nodes.

  </description>

</property>

 

<property>

  <name>mapred.task.tracker.report.address</name>

  <value>127.0.0.1:0</value>

  <description>The interface and port that task tracker server listens on. 

  Since it is only connected to by the tasks, it uses the local interface.

  EXPERT ONLY. Should only be changed if your host does not have the loopback 

  interface.</description>

</property>

 

<property>

  <name>mapred.local.dir</name>

  <value>/home/hadoop/mapred/mapred.local.dir</value>

  <description>The local directory where MapReduce stores intermediate

  data files.  May be a comma-separated list of

  directories on different devices in order to spread disk i/o.

  Directories that do not exist are ignored.

  </description>

</property>

 

<property>

  <name>mapred.system.dir</name>

  <value>/home/hadoop/mapred/system</value>

  <description>The shared directory where MapReduce stores control files.

  </description>

</property>

 

<property>

  <name>mapred.temp.dir</name>

  <value>${hadoop.tmp.dir}/mapred/temp</value>

  <description>A shared directory for temporary files.

  </description>

</property>

 

<property>

  <name>mapred.local.dir.minspacestart</name>

  <value>0</value>

  <description>If the space in mapred.local.dir drops under this, 

  do not ask for more tasks.

  Value in bytes.

  </description>

</property>

 

<property>

  <name>mapred.local.dir.minspacekill</name>

  <value>0</value>

  <description>If the space in mapred.local.dir drops under this, 

    do not ask more tasks until all the current ones have finished and 

    cleaned up. Also, to save the rest of the tasks we have running, 

    kill one of them, to clean up some space. Start with the reduce tasks,

    then go with the ones that have finished the least.

    Value in bytes.

  </description>

</property>

 

<property>

  <name>mapred.tasktracker.expiry.interval</name>

  <value>600000</value>

  <description>Expert: The time-interval, in miliseconds, after which

  a tasktracker is declared 'lost' if it doesn't send heartbeats.

  </description>

</property>

 

<property>

  <name>mapred.tasktracker.instrumentation</name>

  <value>org.apache.hadoop.mapred.TaskTrackerMetricsInst</value>

  <description>Expert: The instrumentation class to associate with each TaskTracker.

  </description>

</property>

 

<property>

  <name>mapred.tasktracker.memory_calculator_plugin</name>

  <value></value>

  <description>

   Name of the class whose instance will be used to query memory information

   on the tasktracker.

 

   The class must be an instance of 

   org.apache.hadoop.util.MemoryCalculatorPlugin. If the value is null, the

   tasktracker attempts to use a class appropriate to the platform. 

   Currently, the only platform supported is Linux.

  </description>

</property>

 

<property>

  <name>mapred.tasktracker.taskmemorymanager.monitoring-interval</name>

  <value>5000</value>

  <description>The interval, in milliseconds, for which the tasktracker waits

   between two cycles of monitoring its tasks' memory usage. Used only if

   tasks' memory management is enabled via mapred.tasktracker.tasks.maxmemory.

   </description>

</property>

 

<property>

  <name>mapred.tasktracker.procfsbasedprocesstree.sleeptime-before-sigkill</name>

  <value>5000</value>

  <description>The time, in milliseconds, the tasktracker waits for sending a

  SIGKILL to a process that has overrun memory limits, after it has been sent

  a SIGTERM. Used only if tasks' memory management is enabled via

  mapred.tasktracker.tasks.maxmemory.</description>

</property>

 

<property>

  <name>mapred.map.tasks</name>

  <value>2</value>

  <description>The default number of map tasks per job.

  Ignored when mapred.job.tracker is "local".  

  </description>

</property>

 

<property>

  <name>mapred.reduce.tasks</name>

  <value>1</value>

  <description>The default number of reduce tasks per job. Typically set to 99%

  of the cluster's reduce capacity, so that if a node fails the reduces can 

  still be executed in a single wave.

  Ignored when mapred.job.tracker is "local".

  </description>

</property>

 

<property>

  <name>mapred.jobtracker.restart.recover</name>

  <value>false</value>

  <description>"true" to enable (job) recovery upon restart,

               "false" to start afresh

  </description>

</property>

 

<property>

  <name>mapred.jobtracker.job.history.block.size</name>

  <value>3145728</value>

  <description>The block size of the job history file. Since the job recovery

               uses job history, its important to dump job history to disk as 

               soon as possible. Note that this is an expert level parameter.

               The default value is set to 3 MB.

  </description>

</property>

 

<property>

  <name>mapred.jobtracker.taskScheduler</name>

  <value>org.apache.hadoop.mapred.JobQueueTaskScheduler</value>

  <description>The class responsible for scheduling the tasks.</description>

</property>

 

<property>

  <name>mapred.jobtracker.taskScheduler.maxRunningTasksPerJob</name>

  <value></value>

  <description>The maximum number of running tasks for a job before

  it gets preempted. No limits if undefined.

  </description>

</property>

 

<property>

  <name>mapred.map.max.attempts</name>

  <value>4</value>

  <description>Expert: The maximum number of attempts per map task.

  In other words, framework will try to execute a map task these many number

  of times before giving up on it.

  </description>

</property>

 

<property>

  <name>mapred.reduce.max.attempts</name>

  <value>4</value>

  <description>Expert: The maximum number of attempts per reduce task.

  In other words, framework will try to execute a reduce task these many number

  of times before giving up on it.

  </description>

</property>

 

<property>

  <name>mapred.reduce.parallel.copies</name>

  <value>5</value>

  <description>The default number of parallel transfers run by reduce

  during the copy(shuffle) phase.

  </description>

</property>

 

<property>

  <name>mapred.reduce.copy.backoff</name>

  <value>300</value>

  <description>The maximum amount of time (in seconds) a reducer spends on 

  fetching one map output before declaring it as failed.

  </description>

</property>

 

<property>

  <name>mapred.task.timeout</name>

  <value>600000</value>

  <description>The number of milliseconds before a task will be

  terminated if it neither reads an input, writes an output, nor

  updates its status string.

  </description>

</property>

 

<property>

  <name>mapred.tasktracker.map.tasks.maximum</name>

  <value>2</value>

  <description>The maximum number of map tasks that will be run

  simultaneously by a task tracker.

  </description>

</property>

 

<property>

  <name>mapred.tasktracker.reduce.tasks.maximum</name>

  <value>2</value>

  <description>The maximum number of reduce tasks that will be run

  simultaneously by a task tracker.

  </description>

</property>

 

<property>

  <name>mapred.jobtracker.completeuserjobs.maximum</name>

  <value>100</value>

  <description>The maximum number of complete jobs per user to keep around 

  before delegating them to the job history.</description>

</property>

 

<property>

  <name>mapred.jobtracker.instrumentation</name>

  <value>org.apache.hadoop.mapred.JobTrackerMetricsInst</value>

  <description>Expert: The instrumentation class to associate with each JobTracker.

  </description>

</property>

 

<property>

  <name>mapred.child.java.opts</name>

  <value>-Xmx200m</value>

  <description>Java opts for the task tracker child processes.  

  The following symbol, if present, will be interpolated: @taskid@ is replaced 

  by current TaskID. Any other occurrences of '@' will go unchanged.

  For example, to enable verbose gc logging to a file named for the taskid in

  /tmp and to set the heap maximum to be a gigabyte, pass a 'value' of:

        -Xmx1024m -verbose:gc -Xloggc:/tmp/@taskid@.gc

 

  The configuration variable mapred.child.ulimit can be used to control the

  maximum virtual memory of the child processes. 

  </description>

</property>

 

<property>

  <name>mapred.child.ulimit</name>

  <value></value>

  <description>The maximum virtual memory, in KB, of a process launched by the 

  Map-Reduce framework. This can be used to control both the Mapper/Reducer 

  tasks and applications using Hadoop Pipes, Hadoop Streaming etc. 

  By default it is left unspecified to let cluster admins control it via 

  limits.conf and other such relevant mechanisms.

 

  Note: mapred.child.ulimit must be greater than or equal to the -Xmx passed to

  JavaVM, else the VM might not start. 

  </description>

</property>

 

<property>

  <name>mapred.child.tmp</name>

  <value>./tmp</value>

  <description> To set the value of tmp directory for map and reduce tasks.

  If the value is an absolute path, it is directly assigned. Otherwise, it is

  prepended with task's working directory. The java tasks are executed with

  option -Djava.io.tmpdir='the absolute path of the tmp dir'. Pipes and

  streaming are set with environment variable,

   TMPDIR='the absolute path of the tmp dir'

  </description>

</property>

 

<property>

  <name>mapred.inmem.merge.threshold</name>

  <value>1000</value>

  <description>The threshold, in terms of the number of files 

  for the in-memory merge process. When we accumulate threshold number of files

  we initiate the in-memory merge and spill to disk. A value of 0 or less than

  0 indicates we want to DON'T have any threshold and instead depend only on

  the ramfs's memory consumption to trigger the merge.

  </description>

</property>

 

<property>

  <name>mapred.job.shuffle.merge.percent</name>

  <value>0.66</value>

  <description>The usage threshold at which an in-memory merge will be

  initiated, expressed as a percentage of the total memory allocated to

  storing in-memory map outputs, as defined by

  mapred.job.shuffle.input.buffer.percent.

  </description>

</property>

 

<property>

  <name>mapred.job.shuffle.input.buffer.percent</name>

  <value>0.70</value>

  <description>The percentage of memory to be allocated from the maximum heap

  size to storing map outputs during the shuffle.

  </description>

</property>

 

<property>

  <name>mapred.job.reduce.input.buffer.percent</name>

  <value>0.0</value>

  <description>The percentage of memory- relative to the maximum heap size- to

  retain map outputs during the reduce. When the shuffle is concluded, any

  remaining map outputs in memory must consume less than this threshold before

  the reduce can begin.

  </description>

</property>

 

<property>

  <name>mapred.map.tasks.speculative.execution</name>

  <value>true</value>

  <description>If true, then multiple instances of some map tasks 

               may be executed in parallel.</description>

</property>

 

<property>

  <name>mapred.reduce.tasks.speculative.execution</name>

  <value>true</value>

  <description>If true, then multiple instances of some reduce tasks 

               may be executed in parallel.</description>

</property>

 

<property>

  <name>mapred.job.reuse.jvm.num.tasks</name>

  <value>1</value>

  <description>How many tasks to run per jvm. If set to -1, there is

  no limit. 

  </description>

</property>

 

<property>

  <name>mapred.min.split.size</name>

  <value>0</value>

  <description>The minimum size chunk that map input should be split

  into.  Note that some file formats may have minimum split sizes that

  take priority over this setting.</description>

</property>

 

<property>

  <name>mapred.jobtracker.maxtasks.per.job</name>

  <value>-1</value>

  <description>The maximum number of tasks for a single job.

  A value of -1 indicates that there is no maximum.  </description>

</property>

 

<property>

  <name>mapred.submit.replication</name>

  <value>10</value>

  <description>The replication level for submitted job files.  This

  should be around the square root of the number of nodes.

  </description>

</property>

 

 

<property>

  <name>mapred.tasktracker.dns.interface</name>

  <value>default</value>

  <description>The name of the Network Interface from which a task

  tracker should report its IP address.

  </description>

 </property>

 

<property>

  <name>mapred.tasktracker.dns.nameserver</name>

  <value>default</value>

  <description>The host name or IP address of the name server (DNS)

  which a TaskTracker should use to determine the host name used by

  the JobTracker for communication and display purposes.

  </description>

 </property>

 

<property>

  <name>tasktracker.http.threads</name>

  <value>40</value>

  <description>The number of worker threads that for the http server. This is

               used for map output fetching

  </description>

</property>

 

<property>

  <name>mapred.task.tracker.http.address</name>

  <value>0.0.0.0:50060</value>

  <description>

    The task tracker http server address and port.

    If the port is 0 then the server will start on a free port.

  </description>

</property>

 

<property>

  <name>keep.failed.task.files</name>

  <value>false</value>

  <description>Should the files for failed tasks be kept. This should only be 

               used on jobs that are failing, because the storage is never

               reclaimed. It also prevents the map outputs from being erased

               from the reduce directory as they are consumed.</description>

</property>

 

 

<!-- 

  <property>

  <name>keep.task.files.pattern</name>

  <value>.*_m_123456_0</value>

  <description>Keep all files from tasks whose task names match the given

               regular expression. Defaults to none.</description>

  </property>

-->

 

<property>

  <name>mapred.output.compress</name>

  <value>false</value>

  <description>Should the job outputs be compressed?

  </description>

</property>

 

<property>

  <name>mapred.output.compression.type</name>

  <value>RECORD</value>

  <description>If the job outputs are to compressed as SequenceFiles, how should

               they be compressed? Should be one of NONE, RECORD or BLOCK.

  </description>

</property>

 

<property>

  <name>mapred.output.compression.codec</name>

  <value>org.apache.hadoop.io.compress.DefaultCodec</value>

  <description>If the job outputs are compressed, how should they be compressed?

  </description>

</property>

 

<property>

  <name>mapred.compress.map.output</name>

  <value>false</value>

  <description>Should the outputs of the maps be compressed before being

               sent across the network. Uses SequenceFile compression.

  </description>

</property>

 

<property>

  <name>mapred.map.output.compression.codec</name>

  <value>org.apache.hadoop.io.compress.DefaultCodec</value>

  <description>If the map outputs are compressed, how should they be 

               compressed?

  </description>

</property>

 

<property>

  <name>map.sort.class</name>

  <value>org.apache.hadoop.util.QuickSort</value>

  <description>The default sort class for sorting keys.

  </description>

</property>

 

<property>

  <name>mapred.userlog.limit.kb</name>

  <value>0</value>

  <description>The maximum size of user-logs of each task in KB. 0 disables the cap.

  </description>

</property>

 

<property>

  <name>mapred.userlog.retain.hours</name>

  <value>24</value>

  <description>The maximum time, in hours, for which the user-logs are to be 

          retained.

  </description>

</property>

 

<property>

  <name>mapred.hosts</name>

  <value></value>

  <description>Names a file that contains the list of nodes that may

  connect to the jobtracker.  If the value is empty, all hosts are

  permitted.</description>

</property>

 

<property>

  <name>mapred.hosts.exclude</name>

  <value></value>

  <description>Names a file that contains the list of hosts that

  should be excluded by the jobtracker.  If the value is empty, no

  hosts are excluded.</description>

</property>

 

<property>

  <name>mapred.max.tracker.blacklists</name>

  <value>4</value>

  <description>The number of blacklists for a taskTracker by various jobs

               after which the task tracker could be blacklisted across

               all jobs. The tracker will be given a tasks later

               (after a day). The tracker will become a healthy

               tracker after a restart.

  </description>

</property> 

 

<property>

  <name>mapred.max.tracker.failures</name>

  <value>4</value>

  <description>The number of task-failures on a tasktracker of a given job 

               after which new tasks of that job aren't assigned to it.

  </description>

</property>

 

<property>

  <name>jobclient.output.filter</name>

  <value>FAILED</value>

  <description>The filter for controlling the output of the task's userlogs sent

               to the console of the JobClient. 

               The permissible options are: NONE, KILLED, FAILED, SUCCEEDED and 

               ALL.

  </description>

</property>

 

  <property>

    <name>mapred.job.tracker.persist.jobstatus.active</name>

    <value>false</value>

    <description>Indicates if persistency of job status information is

      active or not.

    </description>

  </property>

 

  <property>

  <name>mapred.job.tracker.persist.jobstatus.hours</name>

  <value>0</value>

  <description>The number of hours job status information is persisted in DFS.

    The job status information will be available after it drops of the memory

    queue and between jobtracker restarts. With a zero value the job status

    information is not persisted at all in DFS.

  </description>

</property>

 

  <property>

    <name>mapred.job.tracker.persist.jobstatus.dir</name>

    <value>/jobtracker/jobsInfo</value>

    <description>The directory where the job status information is persisted

      in a file system to be available after it drops of the memory queue and

      between jobtracker restarts.

    </description>

  </property>

 

  <property>

    <name>mapred.task.profile</name>

    <value>false</value>

    <description>To set whether the system should collect profiler

     information for some of the tasks in this job? The information is stored

     in the user log directory. The value is "true" if task profiling

     is enabled.</description>

  </property>

 

  <property>

    <name>mapred.task.profile.maps</name>

    <value>0-2</value>

    <description> To set the ranges of map tasks to profile.

    mapred.task.profile has to be set to true for the value to be accounted.

    </description>

  </property>

 

  <property>

    <name>mapred.task.profile.reduces</name>

    <value>0-2</value>

    <description> To set the ranges of reduce tasks to profile.

    mapred.task.profile has to be set to true for the value to be accounted.

    </description>

  </property>

 

  <property>

    <name>mapred.line.input.format.linespermap</name>

    <value>1</value>

    <description> Number of lines per split in NLineInputFormat.

    </description>

  </property>

 

  <property>

    <name>mapred.skip.attempts.to.start.skipping</name>

    <value>2</value>

    <description> The number of Task attempts AFTER which skip mode 

    will be kicked off. When skip mode is kicked off, the 

    tasks reports the range of records which it will process 

    next, to the TaskTracker. So that on failures, TT knows which 

    ones are possibly the bad records. On further executions, 

    those are skipped.

    </description>

  </property>

 

  <property>

    <name>mapred.skip.map.auto.incr.proc.count</name>

    <value>true</value>

    <description> The flag which if set to true, 

    SkipBadRecords.COUNTER_MAP_PROCESSED_RECORDS is incremented 

    by MapRunner after invoking the map function. This value must be set to 

    false for applications which process the records asynchronously 

    or buffer the input records. For example streaming. 

    In such cases applications should increment this counter on their own.

    </description>

  </property>

 

  <property>

    <name>mapred.skip.reduce.auto.incr.proc.count</name>

    <value>true</value>

    <description> The flag which if set to true, 

    SkipBadRecords.COUNTER_REDUCE_PROCESSED_GROUPS is incremented 

    by framework after invoking the reduce function. This value must be set to 

    false for applications which process the records asynchronously 

    or buffer the input records. For example streaming. 

    In such cases applications should increment this counter on their own.

    </description>

  </property>

 

  <property>

    <name>mapred.skip.out.dir</name>

    <value></value>

    <description> If no value is specified here, the skipped records are 

    written to the output directory at _logs/skip.

    User can stop writing skipped records by giving the value "none". 

    </description>

  </property>

 

  <property>

    <name>mapred.skip.map.max.skip.records</name>

    <value>0</value>

    <description> The number of acceptable skip records surrounding the bad 

    record PER bad record in mapper. The number includes the bad record as well.

    To turn the feature of detection/skipping of bad records off, set the 

    value to 0.

    The framework tries to narrow down the skipped range by retrying  

    until this threshold is met OR all attempts get exhausted for this task. 

    Set the value to Long.MAX_VALUE to indicate that framework need not try to 

    narrow down. Whatever records(depends on application) get skipped are 

    acceptable.

    </description>

  </property>

 

  <property>

    <name>mapred.skip.reduce.max.skip.groups</name>

    <value>0</value>

    <description> The number of acceptable skip groups surrounding the bad 

    group PER bad group in reducer. The number includes the bad group as well.

    To turn the feature of detection/skipping of bad groups off, set the 

    value to 0.

    The framework tries to narrow down the skipped range by retrying  

    until this threshold is met OR all attempts get exhausted for this task. 

    Set the value to Long.MAX_VALUE to indicate that framework need not try to 

    narrow down. Whatever groups(depends on application) get skipped are 

    acceptable.

    </description>

  </property>

 

<!-- Job Notification Configuration -->

 

<!--

<property>

 <name>job.end.notification.url</name>

 <value>http://localhost:8080/jobstatus.php?jobId=$jobId&amp;jobStatus=$jobStatus</value>

 <description>Indicates url which will be called on completion of job to inform

              end status of job.

              User can give at most 2 variables with URI : $jobId and $jobStatus.

              If they are present in URI, then they will be replaced by their

              respective values.

</description>

</property>

-->

 

<property>

  <name>job.end.retry.attempts</name>

  <value>0</value>

  <description>Indicates how many times hadoop should attempt to contact the

               notification URL </description>

</property>

 

<property>

  <name>job.end.retry.interval</name>

   <value>30000</value>

   <description>Indicates time in milliseconds between notification URL retry

                calls</description>

</property>

 

<!-- Proxy Configuration -->

<property>

  <name>hadoop.rpc.socket.factory.class.JobSubmissionProtocol</name>

  <value></value>

  <description> SocketFactory to use to connect to a Map/Reduce master

    (JobTracker). If null or empty, then use hadoop.rpc.socket.class.default.

  </description>

</property>

 

<property>

  <name>mapred.task.cache.levels</name>

  <value>2</value>

  <description> This is the max level of the task cache. For example, if

    the level is 2, the tasks cached are at the host level and at the rack

    level.

  </description>

</property>

 

<property>

  <name>mapred.queue.names</name>

  <value>default</value>

  <description> Comma separated list of queues configured for this jobtracker.

    Jobs are added to queues and schedulers can configure different 

    scheduling properties for the various queues. To configure a property 

    for a queue, the name of the queue must match the name specified in this 

    value. Queue properties that are common to all schedulers are configured 

    here with the naming convention, mapred.queue.$QUEUE-NAME.$PROPERTY-NAME,

    for e.g. mapred.queue.default.submit-job-acl.

    The number of queues configured in this parameter could depend on the

    type of scheduler being used, as specified in 

    mapred.jobtracker.taskScheduler. For example, the JobQueueTaskScheduler

    supports only a single queue, which is the default configured here.

    Before adding more queues, ensure that the scheduler you've configured

    supports multiple queues.

  </description>

</property>

 

<property>

  <name>mapred.acls.enabled</name>

  <value>false</value>

  <description> Specifies whether ACLs are enabled, and should be checked

    for various operations.

  </description>

</property>

 

<property>

  <name>mapred.queue.default.acl-submit-job</name>

  <value>*</value>

  <description> Comma separated list of user and group names that are allowed

    to submit jobs to the 'default' queue. The user list and the group list

    are separated by a blank. For e.g. alice,bob group1,group2. 

    If set to the special value '*', it means all users are allowed to 

    submit jobs. 

  </description>

</property>

 

<property>

  <name>mapred.queue.default.acl-administer-jobs</name>

  <value>*</value>

  <description> Comma separated list of user and group names that are allowed

    to delete jobs or modify job's priority for jobs not owned by the current

    user in the 'default' queue. The user list and the group list

    are separated by a blank. For e.g. alice,bob group1,group2. 

    If set to the special value '*', it means all users are allowed to do 

    this operation.

  </description>

</property>

 

<property>

  <name>mapred.job.queue.name</name>

  <value>default</value>

  <description> Queue to which a job is submitted. This must match one of the

    queues defined in mapred.queue.names for the system. Also, the ACL setup

    for the queue must allow the current user to submit a job to the queue.

    Before specifying a queue, ensure that the system is configured with 

    the queue, and access is allowed for submitting jobs to the queue.

  </description>

</property>

 

<property>

  <name>mapred.tasktracker.indexcache.mb</name>

  <value>10</value>

  <description> The maximum memory that a task tracker allows for the 

    index cache that is used when serving map outputs to reducers.

  </description>

</property>

 

<property>

  <name>mapred.merge.recordsBeforeProgress</name>

  <value>10000</value>

  <description> The number of records to process during merge before

   sending a progress notification to the TaskTracker.

  </description>

</property>

 

<property>

  <name>mapred.reduce.slowstart.completed.maps</name>

  <value>0.05</value>

  <description>Fraction of the number of maps in the job which should be 

  complete before reduces are scheduled for the job. 

  </description>

</property>

 

</configuration>