hadoop-default.xml 29 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975976977978979980981982983984985986
  1. <?xml version="1.0"?>
  2. <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
  3. <!-- Do not modify this file directly. Instead, copy entries that you -->
  4. <!-- wish to modify from this file into hadoop-site.xml and change them -->
  5. <!-- there. If hadoop-site.xml does not already exist, create it. -->
  6. <configuration>
  7. <!--- global properties -->
  8. <property>
  9. <name>hadoop.tmp.dir</name>
  10. <value>/tmp/hadoop-${user.name}</value>
  11. <description>A base for other temporary directories.</description>
  12. </property>
  13. <property>
  14. <name>hadoop.native.lib</name>
  15. <value>true</value>
  16. <description>Should native hadoop libraries, if present, be used.</description>
  17. </property>
  18. <!--- logging properties -->
  19. <property>
  20. <name>hadoop.logfile.size</name>
  21. <value>10000000</value>
  22. <description>The max size of each log file</description>
  23. </property>
  24. <property>
  25. <name>hadoop.logfile.count</name>
  26. <value>10</value>
  27. <description>The max number of log files</description>
  28. </property>
  29. <property>
  30. <name>dfs.namenode.logging.level</name>
  31. <value>info</value>
  32. <description>The logging level for dfs namenode. Other values are "dir"(trac
  33. e namespace mutations), "block"(trace block under/over replications and block
  34. creations/deletions), or "all".</description>
  35. </property>
  36. <!-- i/o properties -->
  37. <property>
  38. <name>io.sort.factor</name>
  39. <value>10</value>
  40. <description>The number of streams to merge at once while sorting
  41. files. This determines the number of open file handles.</description>
  42. </property>
  43. <property>
  44. <name>io.sort.mb</name>
  45. <value>100</value>
  46. <description>The total amount of buffer memory to use while sorting
  47. files, in megabytes. By default, gives each merge stream 1MB, which
  48. should minimize seeks.</description>
  49. </property>
  50. <property>
  51. <name>io.file.buffer.size</name>
  52. <value>4096</value>
  53. <description>The size of buffer for use in sequence files.
  54. The size of this buffer should probably be a multiple of hardware
  55. page size (4096 on Intel x86), and it determines how much data is
  56. buffered during read and write operations.</description>
  57. </property>
  58. <property>
  59. <name>io.bytes.per.checksum</name>
  60. <value>512</value>
  61. <description>The number of bytes per checksum. Must not be larger than
  62. io.file.buffer.size.</description>
  63. </property>
  64. <property>
  65. <name>io.skip.checksum.errors</name>
  66. <value>false</value>
  67. <description>If true, when a checksum error is encountered while
  68. reading a sequence file, entries are skipped, instead of throwing an
  69. exception.</description>
  70. </property>
  71. <property>
  72. <name>io.map.index.skip</name>
  73. <value>0</value>
  74. <description>Number of index entries to skip between each entry.
  75. Zero by default. Setting this to values larger than zero can
  76. facilitate opening large map files using less memory.</description>
  77. </property>
  78. <property>
  79. <name>io.compression.codecs</name>
  80. <value>org.apache.hadoop.io.compress.DefaultCodec,org.apache.hadoop.io.compress.GzipCodec</value>
  81. <description>A list of the compression codec classes that can be used
  82. for compression/decompression.</description>
  83. </property>
  84. <!-- file system properties -->
  85. <property>
  86. <name>fs.default.name</name>
  87. <value>file:///</value>
  88. <description>The name of the default file system. A URI whose
  89. scheme and authority determine the FileSystem implementation. The
  90. uri's scheme determines the config property (fs.SCHEME.impl) naming
  91. the FileSystem implementation class. The uri's authority is used to
  92. determine the host, port, etc. for a filesystem.</description>
  93. </property>
  94. <property>
  95. <name>fs.trash.root</name>
  96. <value>${hadoop.tmp.dir}/Trash</value>
  97. <description>The trash directory, used by FsShell's 'rm' command.
  98. </description>
  99. </property>
  100. <property>
  101. <name>fs.trash.interval</name>
  102. <value>0</value>
  103. <description>Number of minutes between trash checkpoints.
  104. If zero, the trash feature is disabled.
  105. </description>
  106. </property>
  107. <property>
  108. <name>fs.file.impl</name>
  109. <value>org.apache.hadoop.fs.LocalFileSystem</value>
  110. <description>The FileSystem for file: uris.</description>
  111. </property>
  112. <property>
  113. <name>fs.hdfs.impl</name>
  114. <value>org.apache.hadoop.dfs.DistributedFileSystem</value>
  115. <description>The FileSystem for hdfs: uris.</description>
  116. </property>
  117. <property>
  118. <name>fs.s3.impl</name>
  119. <value>org.apache.hadoop.fs.s3.S3FileSystem</value>
  120. <description>The FileSystem for s3: uris.</description>
  121. </property>
  122. <property>
  123. <name>fs.kfs.impl</name>
  124. <value>org.apache.hadoop.fs.kfs.KosmosFileSystem</value>
  125. <description>The FileSystem for kfs: uris.</description>
  126. </property>
  127. <property>
  128. <name>fs.hftp.impl</name>
  129. <value>org.apache.hadoop.dfs.HftpFileSystem</value>
  130. </property>
  131. <property>
  132. <name>fs.ramfs.impl</name>
  133. <value>org.apache.hadoop.fs.InMemoryFileSystem</value>
  134. <description>The FileSystem for ramfs: uris.</description>
  135. </property>
  136. <property>
  137. <name>fs.inmemory.size.mb</name>
  138. <value>75</value>
  139. <description>The size of the in-memory filsystem instance in MB</description>
  140. </property>
  141. <property>
  142. <name>fs.checkpoint.dir</name>
  143. <value>${hadoop.tmp.dir}/dfs/namesecondary</value>
  144. <description>Determines where on the local filesystem the DFS secondary
  145. name node should store the temporary images and edits to merge.
  146. </description>
  147. </property>
  148. <property>
  149. <name>fs.checkpoint.period</name>
  150. <value>3600</value>
  151. <description>The number of seconds between two periodic checkpoints.
  152. </description>
  153. </property>
  154. <property>
  155. <name>fs.checkpoint.size</name>
  156. <value>67108864</value>
  157. <description>The size of the current edit log (in bytes) that triggers
  158. a periodic checkpoint even if the fs.checkpoint.period hasn't expired.
  159. </description>
  160. </property>
  161. <property>
  162. <name>dfs.secondary.http.bindAddress</name>
  163. <value>0.0.0.0:50090</value>
  164. <description>
  165. The secondary namenode http server bind address and port.
  166. If the port is 0 then the server will start on a free port.
  167. </description>
  168. </property>
  169. <property>
  170. <name>dfs.datanode.bindAddress</name>
  171. <value>0.0.0.0:50010</value>
  172. <description>
  173. The address where the datanode will listen to.
  174. If the port is 0 then the server will start on a free port.
  175. </description>
  176. </property>
  177. <property>
  178. <name>dfs.datanode.http.bindAddress</name>
  179. <value>0.0.0.0:50075</value>
  180. <description>
  181. The datanode http server bind address and port.
  182. If the port is 0 then the server will start on a free port.
  183. </description>
  184. </property>
  185. <property>
  186. <name>dfs.http.bindAddress</name>
  187. <value>0.0.0.0:50070</value>
  188. <description>
  189. The address and the base port where the dfs namenode web ui will listen on.
  190. If the port is 0 then the server will start on a free port.
  191. </description>
  192. </property>
  193. <property>
  194. <name>dfs.datanode.dns.interface</name>
  195. <value>default</value>
  196. <description>The name of the Network Interface from which a data node should
  197. report its IP address.
  198. </description>
  199. </property>
  200. <property>
  201. <name>dfs.datanode.dns.nameserver</name>
  202. <value>default</value>
  203. <description>The host name or IP address of the name server (DNS)
  204. which a DataNode should use to determine the host name used by the
  205. NameNode for communication and display purposes.
  206. </description>
  207. </property>
  208. <property>
  209. <name>dfs.replication.considerLoad</name>
  210. <value>true</value>
  211. <description>Decide if chooseTarget considers the target's load or not
  212. </description>
  213. </property>
  214. <property>
  215. <name>dfs.default.chunk.view.size</name>
  216. <value>32768</value>
  217. <description>The number of bytes to view for a file on the browser.
  218. </description>
  219. </property>
  220. <property>
  221. <name>dfs.datanode.du.reserved</name>
  222. <value>0</value>
  223. <description>Reserved space in bytes per volume. Always leave this much space free for non dfs use.
  224. </description>
  225. </property>
  226. <property>
  227. <name>dfs.datanode.du.pct</name>
  228. <value>0.98f</value>
  229. <description>When calculating remaining space, only use this percentage of the real available space
  230. </description>
  231. </property>
  232. <property>
  233. <name>dfs.name.dir</name>
  234. <value>${hadoop.tmp.dir}/dfs/name</value>
  235. <description>Determines where on the local filesystem the DFS name node
  236. should store the name table. If this is a comma-delimited list
  237. of directories then the name table is replicated in all of the
  238. directories, for redundancy. </description>
  239. </property>
  240. <property>
  241. <name>dfs.client.buffer.dir</name>
  242. <value>${hadoop.tmp.dir}/dfs/tmp</value>
  243. <description>Determines where on the local filesystem an DFS client
  244. should store its blocks before it sends them to the datanode.
  245. </description>
  246. </property>
  247. <property>
  248. <name>dfs.data.dir</name>
  249. <value>${hadoop.tmp.dir}/dfs/data</value>
  250. <description>Determines where on the local filesystem an DFS data node
  251. should store its blocks. If this is a comma-delimited
  252. list of directories, then data will be stored in all named
  253. directories, typically on different devices.
  254. Directories that do not exist are ignored.
  255. </description>
  256. </property>
  257. <property>
  258. <name>dfs.replication</name>
  259. <value>3</value>
  260. <description>Default block replication.
  261. The actual number of replications can be specified when the file is created.
  262. The default is used if replication is not specified in create time.
  263. </description>
  264. </property>
  265. <property>
  266. <name>dfs.replication.max</name>
  267. <value>512</value>
  268. <description>Maximal block replication.
  269. </description>
  270. </property>
  271. <property>
  272. <name>dfs.replication.min</name>
  273. <value>1</value>
  274. <description>Minimal block replication.
  275. </description>
  276. </property>
  277. <property>
  278. <name>dfs.block.size</name>
  279. <value>67108864</value>
  280. <description>The default block size for new files.</description>
  281. </property>
  282. <property>
  283. <name>dfs.df.interval</name>
  284. <value>60000</value>
  285. <description>Disk usage statistics refresh interval in msec.</description>
  286. </property>
  287. <property>
  288. <name>dfs.client.block.write.retries</name>
  289. <value>3</value>
  290. <description>The number of retries for writing blocks to the data nodes,
  291. before we signal failure to the application.
  292. </description>
  293. </property>
  294. <property>
  295. <name>dfs.blockreport.intervalMsec</name>
  296. <value>3600000</value>
  297. <description>Determines block reporting interval in milliseconds.</description>
  298. </property>
  299. <property>
  300. <name>dfs.heartbeat.interval</name>
  301. <value>3</value>
  302. <description>Determines datanode heartbeat interval in seconds.</description>
  303. </property>
  304. <property>
  305. <name>dfs.namenode.handler.count</name>
  306. <value>10</value>
  307. <description>The number of server threads for the namenode.</description>
  308. </property>
  309. <property>
  310. <name>dfs.safemode.threshold.pct</name>
  311. <value>0.999f</value>
  312. <description>
  313. Specifies the percentage of blocks that should satisfy
  314. the minimal replication requirement defined by dfs.replication.min.
  315. Values less than or equal to 0 mean not to start in safe mode.
  316. Values greater than 1 will make safe mode permanent.
  317. </description>
  318. </property>
  319. <property>
  320. <name>dfs.safemode.extension</name>
  321. <value>30000</value>
  322. <description>
  323. Determines extension of safe mode in milliseconds
  324. after the threshold level is reached.
  325. </description>
  326. </property>
  327. <property>
  328. <name>dfs.network.script</name>
  329. <value></value>
  330. <description>
  331. Specifies a script name that print the network location path
  332. of the current machine.
  333. </description>
  334. </property>
  335. <property>
  336. <name>dfs.hosts</name>
  337. <value></value>
  338. <description>Names a file that contains a list of hosts that are
  339. permitted to connect to the namenode. The full pathname of the file
  340. must be specified. If the value is empty, all hosts are
  341. permitted.</description>
  342. </property>
  343. <property>
  344. <name>dfs.hosts.exclude</name>
  345. <value></value>
  346. <description>Names a file that contains a list of hosts that are
  347. not permitted to connect to the namenode. The full pathname of the
  348. file must be specified. If the value is empty, no hosts are
  349. excluded.</description>
  350. </property>
  351. <property>
  352. <name>fs.s3.block.size</name>
  353. <value>67108864</value>
  354. <description>Block size to use when writing files to S3.</description>
  355. </property>
  356. <property>
  357. <name>fs.s3.buffer.dir</name>
  358. <value>${hadoop.tmp.dir}/s3</value>
  359. <description>Determines where on the local filesystem the S3 filesystem
  360. should store its blocks before it sends them to S3
  361. or after it retrieves them from S3.
  362. </description>
  363. </property>
  364. <property>
  365. <name>fs.s3.maxRetries</name>
  366. <value>4</value>
  367. <description>The maximum number of retries for reading or writing blocks to S3,
  368. before we signal failure to the application.
  369. </description>
  370. </property>
  371. <property>
  372. <name>fs.s3.sleepTimeSeconds</name>
  373. <value>10</value>
  374. <description>The number of seconds to sleep between each S3 retry.
  375. </description>
  376. </property>
  377. <!-- map/reduce properties -->
  378. <property>
  379. <name>mapred.job.tracker</name>
  380. <value>local</value>
  381. <description>The host and port that the MapReduce job tracker runs
  382. at. If "local", then jobs are run in-process as a single map
  383. and reduce task.
  384. </description>
  385. </property>
  386. <property>
  387. <name>mapred.job.tracker.http.bindAddress</name>
  388. <value>0.0.0.0:50030</value>
  389. <description>
  390. The job tracker http server bind address and port.
  391. If the port is 0 then the server will start on a free port.
  392. </description>
  393. </property>
  394. <property>
  395. <name>mapred.job.tracker.handler.count</name>
  396. <value>10</value>
  397. <description>
  398. The number of server threads for the JobTracker. This should be roughly
  399. 4% of the number of tasktracker nodes.
  400. </description>
  401. </property>
  402. <property>
  403. <name>mapred.task.tracker.report.bindAddress</name>
  404. <value>127.0.0.1:0</value>
  405. <description>The interface that task processes use to communicate
  406. with their parent tasktracker process.</description>
  407. </property>
  408. <property>
  409. <name>mapred.local.dir</name>
  410. <value>${hadoop.tmp.dir}/mapred/local</value>
  411. <description>The local directory where MapReduce stores intermediate
  412. data files. May be a comma-separated list of
  413. directories on different devices in order to spread disk i/o.
  414. Directories that do not exist are ignored.
  415. </description>
  416. </property>
  417. <property>
  418. <name>local.cache.size</name>
  419. <value>10737418240</value>
  420. <description>The limit on the size of cache you want to keep, set by default
  421. to 10GB. This will act as a soft limit on the cache directory for out of band data.
  422. </description>
  423. </property>
  424. <property>
  425. <name>mapred.system.dir</name>
  426. <value>${hadoop.tmp.dir}/mapred/system</value>
  427. <description>The shared directory where MapReduce stores control files.
  428. </description>
  429. </property>
  430. <property>
  431. <name>mapred.temp.dir</name>
  432. <value>${hadoop.tmp.dir}/mapred/temp</value>
  433. <description>A shared directory for temporary files.
  434. </description>
  435. </property>
  436. <property>
  437. <name>mapred.local.dir.minspacestart</name>
  438. <value>0</value>
  439. <description>If the space in mapred.local.dir drops under this,
  440. do not ask for more tasks.
  441. Value in bytes.
  442. </description>
  443. </property>
  444. <property>
  445. <name>mapred.local.dir.minspacekill</name>
  446. <value>0</value>
  447. <description>If the space in mapred.local.dir drops under this,
  448. do not ask more tasks until all the current ones have finished and
  449. cleaned up. Also, to save the rest of the tasks we have running,
  450. kill one of them, to clean up some space. Start with the reduce tasks,
  451. then go with the ones that have finished the least.
  452. Value in bytes.
  453. </description>
  454. </property>
  455. <property>
  456. <name>mapred.tasktracker.expiry.interval</name>
  457. <value>600000</value>
  458. <description>Expert: The time-interval, in miliseconds, after which
  459. a tasktracker is declared 'lost' if it doesn't send heartbeats.
  460. </description>
  461. </property>
  462. <property>
  463. <name>mapred.map.tasks</name>
  464. <value>2</value>
  465. <description>The default number of map tasks per job. Typically set
  466. to a prime several times greater than number of available hosts.
  467. Ignored when mapred.job.tracker is "local".
  468. </description>
  469. </property>
  470. <property>
  471. <name>mapred.reduce.tasks</name>
  472. <value>1</value>
  473. <description>The default number of reduce tasks per job. Typically set
  474. to a prime close to the number of available hosts. Ignored when
  475. mapred.job.tracker is "local".
  476. </description>
  477. </property>
  478. <property>
  479. <name>mapred.map.max.attempts</name>
  480. <value>4</value>
  481. <description>Expert: The maximum number of attempts per map task.
  482. In other words, framework will try to execute a map task these many number
  483. of times before giving up on it.
  484. </description>
  485. </property>
  486. <property>
  487. <name>mapred.reduce.max.attempts</name>
  488. <value>4</value>
  489. <description>Expert: The maximum number of attempts per reduce task.
  490. In other words, framework will try to execute a reduce task these many number
  491. of times before giving up on it.
  492. </description>
  493. </property>
  494. <property>
  495. <name>mapred.reduce.parallel.copies</name>
  496. <value>5</value>
  497. <description>The default number of parallel transfers run by reduce
  498. during the copy(shuffle) phase.
  499. </description>
  500. </property>
  501. <property>
  502. <name>mapred.reduce.copy.backoff</name>
  503. <value>300</value>
  504. <description>The maximum amount of time (in seconds) a reducer spends on
  505. fetching one map output before declaring it as failed.
  506. </description>
  507. </property>
  508. <property>
  509. <name>mapred.task.timeout</name>
  510. <value>600000</value>
  511. <description>The number of milliseconds before a task will be
  512. terminated if it neither reads an input, writes an output, nor
  513. updates its status string.
  514. </description>
  515. </property>
  516. <property>
  517. <name>mapred.tasktracker.map.tasks.maximum</name>
  518. <value>2</value>
  519. <description>The maximum number of map tasks that will be run
  520. simultaneously by a task tracker.
  521. </description>
  522. </property>
  523. <property>
  524. <name>mapred.tasktracker.reduce.tasks.maximum</name>
  525. <value>2</value>
  526. <description>The maximum number of reduce tasks that will be run
  527. simultaneously by a task tracker.
  528. </description>
  529. </property>
  530. <property>
  531. <name>mapred.jobtracker.completeuserjobs.maximum</name>
  532. <value>100</value>
  533. <description>The maximum number of complete jobs per user to keep around before delegating them to the job history.
  534. </description>
  535. </property>
  536. <property>
  537. <name>mapred.child.java.opts</name>
  538. <value>-Xmx200m</value>
  539. <description>Java opts for the task tracker child processes. Subsumes
  540. 'mapred.child.heap.size' (If a mapred.child.heap.size value is found
  541. in a configuration, its maximum heap size will be used and a warning
  542. emitted that heap.size has been deprecated). Also, the following symbol,
  543. if present, will be interpolated: @taskid@ is replaced by current TaskID.
  544. Any other occurrences of '@' will go unchanged. For
  545. example, to enable verbose gc logging to a file named for the taskid in
  546. /tmp and to set the heap maximum to be a gigabyte, pass a 'value' of:
  547. -Xmx1024m -verbose:gc -Xloggc:/tmp/@taskid@.gc
  548. </description>
  549. </property>
  550. <property>
  551. <name>mapred.inmem.merge.threshold</name>
  552. <value>1000</value>
  553. <description>The threshold, in terms of the number of files
  554. for the in-memory merge process. When we accumulate threshold number of files
  555. we initiate the in-memory merge and spill to disk. A value of 0 or less than
  556. 0 indicates we want to DON'T have any threshold and instead depend only on
  557. the ramfs's memory consumption to trigger the merge.
  558. </description>
  559. </property>
  560. <property>
  561. <name>mapred.speculative.execution</name>
  562. <value>true</value>
  563. <description>If true, then multiple instances of some map and reduce tasks
  564. may be executed in parallel.</description>
  565. </property>
  566. <property>
  567. <name>mapred.min.split.size</name>
  568. <value>0</value>
  569. <description>The minimum size chunk that map input should be split
  570. into. Note that some file formats may have minimum split sizes that
  571. take priority over this setting.</description>
  572. </property>
  573. <property>
  574. <name>mapred.submit.replication</name>
  575. <value>10</value>
  576. <description>The replication level for submitted job files. This
  577. should be around the square root of the number of nodes.
  578. </description>
  579. </property>
  580. <property>
  581. <name>mapred.tasktracker.dns.interface</name>
  582. <value>default</value>
  583. <description>The name of the Network Interface from which a task
  584. tracker should report its IP address.
  585. </description>
  586. </property>
  587. <property>
  588. <name>mapred.tasktracker.dns.nameserver</name>
  589. <value>default</value>
  590. <description>The host name or IP address of the name server (DNS)
  591. which a TaskTracker should use to determine the host name used by
  592. the JobTracker for communication and display purposes.
  593. </description>
  594. </property>
  595. <property>
  596. <name>tasktracker.http.threads</name>
  597. <value>40</value>
  598. <description>The number of worker threads that for the http server. This is
  599. used for map output fetching
  600. </description>
  601. </property>
  602. <property>
  603. <name>mapred.task.tracker.http.bindAddress</name>
  604. <value>0.0.0.0:50060</value>
  605. <description>
  606. The task tracker http server bind address and port.
  607. If the port is 0 then the server will start on a free port.
  608. </description>
  609. </property>
  610. <property>
  611. <name>keep.failed.task.files</name>
  612. <value>false</value>
  613. <description>Should the files for failed tasks be kept. This should only be
  614. used on jobs that are failing, because the storage is never
  615. reclaimed. It also prevents the map outputs from being erased
  616. from the reduce directory as they are consumed.</description>
  617. </property>
  618. <!--
  619. <property>
  620. <name>keep.task.files.pattern</name>
  621. <value>.*_m_123456_0</value>
  622. <description>Keep all files from tasks whose task names match the given
  623. regular expression. Defaults to none.</description>
  624. </property>
  625. -->
  626. <property>
  627. <name>mapred.output.compress</name>
  628. <value>false</value>
  629. <description>Should the job outputs be compressed?
  630. </description>
  631. </property>
  632. <property>
  633. <name>mapred.output.compression.type</name>
  634. <value>RECORD</value>
  635. <description>If the job outputs are to compressed as SequenceFiles, how should
  636. they be compressed? Should be one of NONE, RECORD or BLOCK.
  637. </description>
  638. </property>
  639. <property>
  640. <name>mapred.output.compression.codec</name>
  641. <value>org.apache.hadoop.io.compress.DefaultCodec</value>
  642. <description>If the job outputs are compressed, how should they be compressed?
  643. </description>
  644. </property>
  645. <property>
  646. <name>mapred.compress.map.output</name>
  647. <value>false</value>
  648. <description>Should the outputs of the maps be compressed before being
  649. sent across the network. Uses SequenceFile compression.
  650. </description>
  651. </property>
  652. <property>
  653. <name>mapred.map.output.compression.type</name>
  654. <value>RECORD</value>
  655. <description>If the map outputs are to compressed, how should they
  656. be compressed? Should be one of NONE, RECORD or BLOCK.
  657. </description>
  658. </property>
  659. <property>
  660. <name>mapred.map.output.compression.codec</name>
  661. <value>org.apache.hadoop.io.compress.DefaultCodec</value>
  662. <description>If the map outputs are compressed, how should they be
  663. compressed?
  664. </description>
  665. </property>
  666. <property>
  667. <name>io.seqfile.compress.blocksize</name>
  668. <value>1000000</value>
  669. <description>The minimum block size for compression in block compressed
  670. SequenceFiles.
  671. </description>
  672. </property>
  673. <property>
  674. <name>io.seqfile.lazydecompress</name>
  675. <value>true</value>
  676. <description>Should values of block-compressed SequenceFiles be decompressed
  677. only when necessary.
  678. </description>
  679. </property>
  680. <property>
  681. <name>io.seqfile.sorter.recordlimit</name>
  682. <value>1000000</value>
  683. <description>The limit on number of records to be kept in memory in a spill
  684. in SequenceFiles.Sorter
  685. </description>
  686. </property>
  687. <property>
  688. <name>io.seqfile.compression.type</name>
  689. <value>RECORD</value>
  690. <description>The default compression type for SequenceFile.Writer.
  691. </description>
  692. </property>
  693. <property>
  694. <name>map.sort.class</name>
  695. <value>org.apache.hadoop.mapred.MergeSorter</value>
  696. <description>The default sort class for sorting keys.
  697. </description>
  698. </property>
  699. <property>
  700. <name>mapred.userlog.limit.kb</name>
  701. <value>0</value>
  702. <description>The maximum size of user-logs of each task in KB. 0 disables the cap.
  703. </description>
  704. </property>
  705. <property>
  706. <name>mapred.userlog.retain.hours</name>
  707. <value>24</value>
  708. <description>The maximum time, in hours, for which the user-logs are to be
  709. retained.
  710. </description>
  711. </property>
  712. <property>
  713. <name>mapred.hosts</name>
  714. <value></value>
  715. <description>Names a file that contains the list of nodes that may
  716. connect to the jobtracker. If the value is empty, all hosts are
  717. permitted.</description>
  718. </property>
  719. <property>
  720. <name>mapred.hosts.exclude</name>
  721. <value></value>
  722. <description>Names a file that contains the list of hosts that
  723. should be excluded by the jobtracker. If the value is empty, no
  724. hosts are excluded.</description>
  725. </property>
  726. <property>
  727. <name>mapred.max.tracker.failures</name>
  728. <value>4</value>
  729. <description>The number of task-failures on a tasktracker of a given job
  730. after which new tasks of that job aren't assigned to it.
  731. </description>
  732. </property>
  733. <property>
  734. <name>jobclient.output.filter</name>
  735. <value>FAILED</value>
  736. <description>The filter for controlling the output of the task's userlogs sent
  737. to the console of the JobClient.
  738. The permissible options are: NONE, KILLED, FAILED, SUCCEEDED and
  739. ALL.
  740. </description>
  741. </property>
  742. <!-- ipc properties -->
  743. <property>
  744. <name>ipc.client.timeout</name>
  745. <value>60000</value>
  746. <description>Defines the timeout for IPC calls in milliseconds.</description>
  747. </property>
  748. <property>
  749. <name>ipc.client.idlethreshold</name>
  750. <value>4000</value>
  751. <description>Defines the threshold number of connections after which
  752. connections will be inspected for idleness.
  753. </description>
  754. </property>
  755. <property>
  756. <name>ipc.client.maxidletime</name>
  757. <value>120000</value>
  758. <description>Defines the maximum idle time for a connected client after
  759. which it may be disconnected.
  760. </description>
  761. </property>
  762. <property>
  763. <name>ipc.client.kill.max</name>
  764. <value>10</value>
  765. <description>Defines the maximum number of clients to disconnect in one go.
  766. </description>
  767. </property>
  768. <property>
  769. <name>ipc.client.connection.maxidletime</name>
  770. <value>1000</value>
  771. <description>The maximum time after which a client will bring down the
  772. connection to the server.
  773. </description>
  774. </property>
  775. <property>
  776. <name>ipc.client.connect.max.retries</name>
  777. <value>10</value>
  778. <description>Indicates the number of retries a client will make to establish
  779. a server connection.
  780. </description>
  781. </property>
  782. <property>
  783. <name>ipc.server.listen.queue.size</name>
  784. <value>128</value>
  785. <description>Indicates the length of the listen queue for servers accepting
  786. client connections.
  787. </description>
  788. </property>
  789. <!-- Job Notification Configuration -->
  790. <!--
  791. <property>
  792. <name>job.end.notification.url</name>
  793. <value>http://localhost:8080/jobstatus.php?jobId=$jobId&amp;jobStatus=$jobStatus</value>
  794. <description>Indicates url which will be called on completion of job to inform
  795. end status of job.
  796. User can give at most 2 variables with URI : $jobId and $jobStatus.
  797. If they are present in URI, then they will be replaced by their
  798. respective values.
  799. </description>
  800. </property>
  801. -->
  802. <property>
  803. <name>job.end.retry.attempts</name>
  804. <value>0</value>
  805. <description>Indicates how many times hadoop should attempt to contact the
  806. notification URL </description>
  807. </property>
  808. <property>
  809. <name>job.end.retry.interval</name>
  810. <value>30000</value>
  811. <description>Indicates time in milliseconds between notification URL retry
  812. calls</description>
  813. </property>
  814. <!-- Web Interface Configuration -->
  815. <property>
  816. <name>webinterface.private.actions</name>
  817. <value>false</value>
  818. <description> If set to true, the web interfaces of JT and NN may contain
  819. actions, such as kill job, delete file, etc., that should
  820. not be exposed to public. Enable this option if the interfaces
  821. are only reachable by those who have the right authorization.
  822. </description>
  823. </property>
  824. <!-- Proxy Configuration -->
  825. <property>
  826. <name>hadoop.rpc.socket.factory.class.default</name>
  827. <value>org.apache.hadoop.net.StandardSocketFactory</value>
  828. <description> Default SocketFactory to use. This parameter is expected to be
  829. formatted as "package.FactoryClassName".
  830. </description>
  831. </property>
  832. <property>
  833. <name>hadoop.rpc.socket.factory.class.ClientProtocol</name>
  834. <value></value>
  835. <description> SocketFactory to use to connect to a DFS. If null or empty, use
  836. hadoop.rpc.socket.class.default. This socket factory is also used by
  837. DFSClient to create sockets to DataNodes.
  838. </description>
  839. </property>
  840. <property>
  841. <name>hadoop.rpc.socket.factory.class.JobSubmissionProtocol</name>
  842. <value></value>
  843. <description> SocketFactory to use to connect to a Map/Reduce master
  844. (JobTracker). If null or empty, then use hadoop.rpc.socket.class.default.
  845. </description>
  846. </property>
  847. <property>
  848. <name>hadoop.socks.server</name>
  849. <value></value>
  850. <description> Address (host:port) of the SOCKS server to be used by the
  851. SocksSocketFactory.
  852. </description>
  853. </property>
  854. </configuration>