mapred-default.xml 55 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574157515761577157815791580158115821583158415851586158715881589159015911592159315941595159615971598
  1. <?xml version="1.0"?>
  2. <!--
  3. Licensed to the Apache Software Foundation (ASF) under one or more
  4. contributor license agreements. See the NOTICE file distributed with
  5. this work for additional information regarding copyright ownership.
  6. The ASF licenses this file to You under the Apache License, Version 2.0
  7. (the "License"); you may not use this file except in compliance with
  8. the License. You may obtain a copy of the License at
  9. http://www.apache.org/licenses/LICENSE-2.0
  10. Unless required by applicable law or agreed to in writing, software
  11. distributed under the License is distributed on an "AS IS" BASIS,
  12. WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  13. See the License for the specific language governing permissions and
  14. limitations under the License.
  15. -->
  16. <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
  17. <!-- Do not modify this file directly. Instead, copy entries that you -->
  18. <!-- wish to modify from this file into mapred-site.xml and change them -->
  19. <!-- there. If mapred-site.xml does not already exist, create it. -->
  20. <configuration>
  21. <property>
  22. <name>mapreduce.jobtracker.jobhistory.location</name>
  23. <value></value>
  24. <description> If job tracker is static the history files are stored
  25. in this single well known place. If No value is set here, by default,
  26. it is in the local file system at ${hadoop.log.dir}/history.
  27. </description>
  28. </property>
  29. <property>
  30. <name>mapreduce.jobtracker.jobhistory.task.numberprogresssplits</name>
  31. <value>12</value>
  32. <description> Every task attempt progresses from 0.0 to 1.0 [unless
  33. it fails or is killed]. We record, for each task attempt, certain
  34. statistics over each twelfth of the progress range. You can change
  35. the number of intervals we divide the entire range of progress into
  36. by setting this property. Higher values give more precision to the
  37. recorded data, but costs more memory in the job tracker at runtime.
  38. Each increment in this attribute costs 16 bytes per running task.
  39. </description>
  40. </property>
  41. <property>
  42. <name>mapreduce.job.userhistorylocation</name>
  43. <value></value>
  44. <description> User can specify a location to store the history files of
  45. a particular job. If nothing is specified, the logs are stored in
  46. output directory. The files are stored in "_logs/history/" in the directory.
  47. User can stop logging by giving the value "none".
  48. </description>
  49. </property>
  50. <property>
  51. <name>mapreduce.jobtracker.jobhistory.completed.location</name>
  52. <value></value>
  53. <description> The completed job history files are stored at this single well
  54. known location. If nothing is specified, the files are stored at
  55. ${mapreduce.jobtracker.jobhistory.location}/done.
  56. </description>
  57. </property>
  58. <property>
  59. <name>mapreduce.job.committer.setup.cleanup.needed</name>
  60. <value>true</value>
  61. <description> true, if job needs job-setup and job-cleanup.
  62. false, otherwise
  63. </description>
  64. </property>
  65. <!-- i/o properties -->
  66. <property>
  67. <name>mapreduce.task.io.sort.factor</name>
  68. <value>10</value>
  69. <description>The number of streams to merge at once while sorting
  70. files. This determines the number of open file handles.</description>
  71. </property>
  72. <property>
  73. <name>mapreduce.task.io.sort.mb</name>
  74. <value>100</value>
  75. <description>The total amount of buffer memory to use while sorting
  76. files, in megabytes. By default, gives each merge stream 1MB, which
  77. should minimize seeks.</description>
  78. </property>
  79. <property>
  80. <name>mapreduce.map.sort.spill.percent</name>
  81. <value>0.80</value>
  82. <description>The soft limit in the serialization buffer. Once reached, a
  83. thread will begin to spill the contents to disk in the background. Note that
  84. collection will not block if this threshold is exceeded while a spill is
  85. already in progress, so spills may be larger than this threshold when it is
  86. set to less than .5</description>
  87. </property>
  88. <property>
  89. <name>mapreduce.jobtracker.address</name>
  90. <value>local</value>
  91. <description>The host and port that the MapReduce job tracker runs
  92. at. If "local", then jobs are run in-process as a single map
  93. and reduce task.
  94. </description>
  95. </property>
  96. <property>
  97. <name>mapreduce.local.clientfactory.class.name</name>
  98. <value>org.apache.hadoop.mapred.LocalClientFactory</value>
  99. <description>This the client factory that is responsible for
  100. creating local job runner client</description>
  101. </property>
  102. <property>
  103. <name>mapreduce.jobtracker.http.address</name>
  104. <value>0.0.0.0:50030</value>
  105. <description>
  106. The job tracker http server address and port the server will listen on.
  107. If the port is 0 then the server will start on a free port.
  108. </description>
  109. </property>
  110. <property>
  111. <name>mapreduce.jobtracker.handler.count</name>
  112. <value>10</value>
  113. <description>
  114. The number of server threads for the JobTracker. This should be roughly
  115. 4% of the number of tasktracker nodes.
  116. </description>
  117. </property>
  118. <property>
  119. <name>mapreduce.tasktracker.report.address</name>
  120. <value>127.0.0.1:0</value>
  121. <description>The interface and port that task tracker server listens on.
  122. Since it is only connected to by the tasks, it uses the local interface.
  123. EXPERT ONLY. Should only be changed if your host does not have the loopback
  124. interface.</description>
  125. </property>
  126. <property>
  127. <name>mapreduce.cluster.local.dir</name>
  128. <value>${hadoop.tmp.dir}/mapred/local</value>
  129. <description>The local directory where MapReduce stores intermediate
  130. data files. May be a comma-separated list of
  131. directories on different devices in order to spread disk i/o.
  132. Directories that do not exist are ignored.
  133. </description>
  134. </property>
  135. <property>
  136. <name>mapreduce.jobtracker.system.dir</name>
  137. <value>${hadoop.tmp.dir}/mapred/system</value>
  138. <description>The directory where MapReduce stores control files.
  139. </description>
  140. </property>
  141. <property>
  142. <name>mapreduce.jobtracker.staging.root.dir</name>
  143. <value>${hadoop.tmp.dir}/mapred/staging</value>
  144. <description>The root of the staging area for users' job files
  145. In practice, this should be the directory where users' home
  146. directories are located (usually /user)
  147. </description>
  148. </property>
  149. <property>
  150. <name>mapreduce.cluster.temp.dir</name>
  151. <value>${hadoop.tmp.dir}/mapred/temp</value>
  152. <description>A shared directory for temporary files.
  153. </description>
  154. </property>
  155. <property>
  156. <name>mapreduce.tasktracker.local.dir.minspacestart</name>
  157. <value>0</value>
  158. <description>If the space in mapreduce.cluster.local.dir drops under this,
  159. do not ask for more tasks.
  160. Value in bytes.
  161. </description>
  162. </property>
  163. <property>
  164. <name>mapreduce.tasktracker.local.dir.minspacekill</name>
  165. <value>0</value>
  166. <description>If the space in mapreduce.cluster.local.dir drops under this,
  167. do not ask more tasks until all the current ones have finished and
  168. cleaned up. Also, to save the rest of the tasks we have running,
  169. kill one of them, to clean up some space. Start with the reduce tasks,
  170. then go with the ones that have finished the least.
  171. Value in bytes.
  172. </description>
  173. </property>
  174. <property>
  175. <name>mapreduce.jobtracker.expire.trackers.interval</name>
  176. <value>600000</value>
  177. <description>Expert: The time-interval, in miliseconds, after which
  178. a tasktracker is declared 'lost' if it doesn't send heartbeats.
  179. </description>
  180. </property>
  181. <property>
  182. <name>mapreduce.tasktracker.instrumentation</name>
  183. <value>org.apache.hadoop.mapred.TaskTrackerMetricsInst</value>
  184. <description>Expert: The instrumentation class to associate with each TaskTracker.
  185. </description>
  186. </property>
  187. <property>
  188. <name>mapreduce.tasktracker.resourcecalculatorplugin</name>
  189. <value></value>
  190. <description>
  191. Name of the class whose instance will be used to query resource information
  192. on the tasktracker.
  193. The class must be an instance of
  194. org.apache.hadoop.util.ResourceCalculatorPlugin. If the value is null, the
  195. tasktracker attempts to use a class appropriate to the platform.
  196. Currently, the only platform supported is Linux.
  197. </description>
  198. </property>
  199. <property>
  200. <name>mapreduce.tasktracker.taskmemorymanager.monitoringinterval</name>
  201. <value>5000</value>
  202. <description>The interval, in milliseconds, for which the tasktracker waits
  203. between two cycles of monitoring its tasks' memory usage. Used only if
  204. tasks' memory management is enabled via mapred.tasktracker.tasks.maxmemory.
  205. </description>
  206. </property>
  207. <property>
  208. <name>mapreduce.tasktracker.tasks.sleeptimebeforesigkill</name>
  209. <value>5000</value>
  210. <description>The time, in milliseconds, the tasktracker waits for sending a
  211. SIGKILL to a task, after it has been sent a SIGTERM. This is currently
  212. not used on WINDOWS where tasks are just sent a SIGTERM.
  213. </description>
  214. </property>
  215. <property>
  216. <name>mapreduce.job.maps</name>
  217. <value>2</value>
  218. <description>The default number of map tasks per job.
  219. Ignored when mapreduce.jobtracker.address is "local".
  220. </description>
  221. </property>
  222. <property>
  223. <name>mapreduce.job.reduces</name>
  224. <value>1</value>
  225. <description>The default number of reduce tasks per job. Typically set to 99%
  226. of the cluster's reduce capacity, so that if a node fails the reduces can
  227. still be executed in a single wave.
  228. Ignored when mapreduce.jobtracker.address is "local".
  229. </description>
  230. </property>
  231. <property>
  232. <name>mapreduce.jobtracker.restart.recover</name>
  233. <value>false</value>
  234. <description>"true" to enable (job) recovery upon restart,
  235. "false" to start afresh
  236. </description>
  237. </property>
  238. <property>
  239. <name>mapreduce.jobtracker.jobhistory.block.size</name>
  240. <value>3145728</value>
  241. <description>The block size of the job history file. Since the job recovery
  242. uses job history, its important to dump job history to disk as
  243. soon as possible. Note that this is an expert level parameter.
  244. The default value is set to 3 MB.
  245. </description>
  246. </property>
  247. <property>
  248. <name>mapreduce.jobtracker.taskscheduler</name>
  249. <value>org.apache.hadoop.mapred.JobQueueTaskScheduler</value>
  250. <description>The class responsible for scheduling the tasks.</description>
  251. </property>
  252. <property>
  253. <name>mapreduce.job.split.metainfo.maxsize</name>
  254. <value>10000000</value>
  255. <description>The maximum permissible size of the split metainfo file.
  256. The JobTracker won't attempt to read split metainfo files bigger than
  257. the configured value.
  258. No limits if set to -1.
  259. </description>
  260. </property>
  261. <property>
  262. <name>mapreduce.jobtracker.taskscheduler.maxrunningtasks.perjob</name>
  263. <value></value>
  264. <description>The maximum number of running tasks for a job before
  265. it gets preempted. No limits if undefined.
  266. </description>
  267. </property>
  268. <property>
  269. <name>mapreduce.map.maxattempts</name>
  270. <value>4</value>
  271. <description>Expert: The maximum number of attempts per map task.
  272. In other words, framework will try to execute a map task these many number
  273. of times before giving up on it.
  274. </description>
  275. </property>
  276. <property>
  277. <name>mapreduce.reduce.maxattempts</name>
  278. <value>4</value>
  279. <description>Expert: The maximum number of attempts per reduce task.
  280. In other words, framework will try to execute a reduce task these many number
  281. of times before giving up on it.
  282. </description>
  283. </property>
  284. <property>
  285. <name>mapreduce.reduce.shuffle.retry-delay.max.ms</name>
  286. <value>60000</value>
  287. <description>The maximum number of ms the reducer will delay before retrying
  288. to download map data.
  289. </description>
  290. </property>
  291. <property>
  292. <name>mapreduce.reduce.shuffle.parallelcopies</name>
  293. <value>5</value>
  294. <description>The default number of parallel transfers run by reduce
  295. during the copy(shuffle) phase.
  296. </description>
  297. </property>
  298. <property>
  299. <name>mapreduce.reduce.shuffle.connect.timeout</name>
  300. <value>180000</value>
  301. <description>Expert: The maximum amount of time (in milli seconds) reduce
  302. task spends in trying to connect to a tasktracker for getting map output.
  303. </description>
  304. </property>
  305. <property>
  306. <name>mapreduce.reduce.shuffle.read.timeout</name>
  307. <value>180000</value>
  308. <description>Expert: The maximum amount of time (in milli seconds) reduce
  309. task waits for map output data to be available for reading after obtaining
  310. connection.
  311. </description>
  312. </property>
  313. <property>
  314. <name>mapreduce.task.timeout</name>
  315. <value>600000</value>
  316. <description>The number of milliseconds before a task will be
  317. terminated if it neither reads an input, writes an output, nor
  318. updates its status string. A value of 0 disables the timeout.
  319. </description>
  320. </property>
  321. <property>
  322. <name>mapreduce.tasktracker.map.tasks.maximum</name>
  323. <value>2</value>
  324. <description>The maximum number of map tasks that will be run
  325. simultaneously by a task tracker.
  326. </description>
  327. </property>
  328. <property>
  329. <name>mapreduce.tasktracker.reduce.tasks.maximum</name>
  330. <value>2</value>
  331. <description>The maximum number of reduce tasks that will be run
  332. simultaneously by a task tracker.
  333. </description>
  334. </property>
  335. <property>
  336. <name>mapreduce.jobtracker.retiredjobs.cache.size</name>
  337. <value>1000</value>
  338. <description>The number of retired job status to keep in the cache.
  339. </description>
  340. </property>
  341. <property>
  342. <name>mapreduce.tasktracker.outofband.heartbeat</name>
  343. <value>false</value>
  344. <description>Expert: Set this to true to let the tasktracker send an
  345. out-of-band heartbeat on task-completion for better latency.
  346. </description>
  347. </property>
  348. <property>
  349. <name>mapreduce.jobtracker.jobhistory.lru.cache.size</name>
  350. <value>5</value>
  351. <description>The number of job history files loaded in memory. The jobs are
  352. loaded when they are first accessed. The cache is cleared based on LRU.
  353. </description>
  354. </property>
  355. <property>
  356. <name>mapreduce.jobtracker.instrumentation</name>
  357. <value>org.apache.hadoop.mapred.JobTrackerMetricsInst</value>
  358. <description>Expert: The instrumentation class to associate with each JobTracker.
  359. </description>
  360. </property>
  361. <property>
  362. <name>mapred.child.java.opts</name>
  363. <value>-Xmx200m</value>
  364. <description>Java opts for the task tracker child processes.
  365. The following symbol, if present, will be interpolated: @taskid@ is replaced
  366. by current TaskID. Any other occurrences of '@' will go unchanged.
  367. For example, to enable verbose gc logging to a file named for the taskid in
  368. /tmp and to set the heap maximum to be a gigabyte, pass a 'value' of:
  369. -Xmx1024m -verbose:gc -Xloggc:/tmp/@taskid@.gc
  370. Usage of -Djava.library.path can cause programs to no longer function if
  371. hadoop native libraries are used. These values should instead be set as part
  372. of LD_LIBRARY_PATH in the map / reduce JVM env using the mapreduce.map.env and
  373. mapreduce.reduce.env config settings.
  374. </description>
  375. </property>
  376. <property>
  377. <name>mapred.child.env</name>
  378. <value></value>
  379. <description>User added environment variables for the task tracker child
  380. processes. Example :
  381. 1) A=foo This will set the env variable A to foo
  382. 2) B=$B:c This is inherit nodemanager's B env variable on Unix.
  383. 3) B=%B%;c This is inherit nodemanager's B env variable on Windows.
  384. </description>
  385. </property>
  386. <property>
  387. <name>mapreduce.admin.user.env</name>
  388. <value>LD_LIBRARY_PATH=$HADOOP_COMMON_HOME/lib/native</value>
  389. <description>Expert: Additional execution environment entries for
  390. map and reduce task processes. This is not an additive property.
  391. You must preserve the original value if you want your map and
  392. reduce tasks to have access to native libraries (compression, etc).
  393. </description>
  394. </property>
  395. <property>
  396. <name>mapreduce.task.tmp.dir</name>
  397. <value>./tmp</value>
  398. <description> To set the value of tmp directory for map and reduce tasks.
  399. If the value is an absolute path, it is directly assigned. Otherwise, it is
  400. prepended with task's working directory. The java tasks are executed with
  401. option -Djava.io.tmpdir='the absolute path of the tmp dir'. Pipes and
  402. streaming are set with environment variable,
  403. TMPDIR='the absolute path of the tmp dir'
  404. </description>
  405. </property>
  406. <property>
  407. <name>mapreduce.map.log.level</name>
  408. <value>INFO</value>
  409. <description>The logging level for the map task. The allowed levels are:
  410. OFF, FATAL, ERROR, WARN, INFO, DEBUG, TRACE and ALL.
  411. </description>
  412. </property>
  413. <property>
  414. <name>mapreduce.reduce.log.level</name>
  415. <value>INFO</value>
  416. <description>The logging level for the reduce task. The allowed levels are:
  417. OFF, FATAL, ERROR, WARN, INFO, DEBUG, TRACE and ALL.
  418. </description>
  419. </property>
  420. <property>
  421. <name>mapreduce.map.cpu.vcores</name>
  422. <value>1</value>
  423. <description>
  424. The number of virtual cores required for each map task.
  425. </description>
  426. </property>
  427. <property>
  428. <name>mapreduce.reduce.cpu.vcores</name>
  429. <value>1</value>
  430. <description>
  431. The number of virtual cores required for each reduce task.
  432. </description>
  433. </property>
  434. <property>
  435. <name>mapreduce.reduce.merge.inmem.threshold</name>
  436. <value>1000</value>
  437. <description>The threshold, in terms of the number of files
  438. for the in-memory merge process. When we accumulate threshold number of files
  439. we initiate the in-memory merge and spill to disk. A value of 0 or less than
  440. 0 indicates we want to DON'T have any threshold and instead depend only on
  441. the ramfs's memory consumption to trigger the merge.
  442. </description>
  443. </property>
  444. <property>
  445. <name>mapreduce.reduce.shuffle.merge.percent</name>
  446. <value>0.66</value>
  447. <description>The usage threshold at which an in-memory merge will be
  448. initiated, expressed as a percentage of the total memory allocated to
  449. storing in-memory map outputs, as defined by
  450. mapreduce.reduce.shuffle.input.buffer.percent.
  451. </description>
  452. </property>
  453. <property>
  454. <name>mapreduce.reduce.shuffle.input.buffer.percent</name>
  455. <value>0.70</value>
  456. <description>The percentage of memory to be allocated from the maximum heap
  457. size to storing map outputs during the shuffle.
  458. </description>
  459. </property>
  460. <property>
  461. <name>mapreduce.reduce.input.buffer.percent</name>
  462. <value>0.0</value>
  463. <description>The percentage of memory- relative to the maximum heap size- to
  464. retain map outputs during the reduce. When the shuffle is concluded, any
  465. remaining map outputs in memory must consume less than this threshold before
  466. the reduce can begin.
  467. </description>
  468. </property>
  469. <property>
  470. <name>mapreduce.reduce.shuffle.memory.limit.percent</name>
  471. <value>0.25</value>
  472. <description>Expert: Maximum percentage of the in-memory limit that a
  473. single shuffle can consume</description>
  474. </property>
  475. <property>
  476. <name>mapreduce.shuffle.ssl.enabled</name>
  477. <value>false</value>
  478. <description>
  479. Whether to use SSL for for the Shuffle HTTP endpoints.
  480. </description>
  481. </property>
  482. <property>
  483. <name>mapreduce.shuffle.ssl.file.buffer.size</name>
  484. <value>65536</value>
  485. <description>Buffer size for reading spills from file when using SSL.
  486. </description>
  487. </property>
  488. <property>
  489. <name>mapreduce.shuffle.max.connections</name>
  490. <value>0</value>
  491. <description>Max allowed connections for the shuffle. Set to 0 (zero)
  492. to indicate no limit on the number of connections.
  493. </description>
  494. </property>
  495. <property>
  496. <name>mapreduce.reduce.markreset.buffer.percent</name>
  497. <value>0.0</value>
  498. <description>The percentage of memory -relative to the maximum heap size- to
  499. be used for caching values when using the mark-reset functionality.
  500. </description>
  501. </property>
  502. <property>
  503. <name>mapreduce.map.speculative</name>
  504. <value>true</value>
  505. <description>If true, then multiple instances of some map tasks
  506. may be executed in parallel.</description>
  507. </property>
  508. <property>
  509. <name>mapreduce.reduce.speculative</name>
  510. <value>true</value>
  511. <description>If true, then multiple instances of some reduce tasks
  512. may be executed in parallel.</description>
  513. </property>
  514. <property>
  515. <name>mapreduce.job.speculative.speculativecap</name>
  516. <value>0.1</value>
  517. <description>The max percent (0-1) of running tasks that
  518. can be speculatively re-executed at any time.</description>
  519. </property>
  520. <property>
  521. <name>mapreduce.job.map.output.collector.class</name>
  522. <value>org.apache.hadoop.mapred.MapTask$MapOutputBuffer</value>
  523. <description>
  524. It defines the MapOutputCollector implementation to use.
  525. </description>
  526. </property>
  527. <property>
  528. <name>mapreduce.job.speculative.slowtaskthreshold</name>
  529. <value>1.0</value>The number of standard deviations by which a task's
  530. ave progress-rates must be lower than the average of all running tasks'
  531. for the task to be considered too slow.
  532. <description>
  533. </description>
  534. </property>
  535. <property>
  536. <name>mapreduce.job.speculative.slownodethreshold</name>
  537. <value>1.0</value>
  538. <description>The number of standard deviations by which a Task
  539. Tracker's ave map and reduce progress-rates (finishTime-dispatchTime)
  540. must be lower than the average of all successful map/reduce task's for
  541. the TT to be considered too slow to give a speculative task to.
  542. </description>
  543. </property>
  544. <property>
  545. <name>mapreduce.job.jvm.numtasks</name>
  546. <value>1</value>
  547. <description>How many tasks to run per jvm. If set to -1, there is
  548. no limit.
  549. </description>
  550. </property>
  551. <property>
  552. <name>mapreduce.job.ubertask.enable</name>
  553. <value>false</value>
  554. <description>Whether to enable the small-jobs "ubertask" optimization,
  555. which runs "sufficiently small" jobs sequentially within a single JVM.
  556. "Small" is defined by the following maxmaps, maxreduces, and maxbytes
  557. settings. Users may override this value.
  558. </description>
  559. </property>
  560. <property>
  561. <name>mapreduce.job.ubertask.maxmaps</name>
  562. <value>9</value>
  563. <description>Threshold for number of maps, beyond which job is considered
  564. too big for the ubertasking optimization. Users may override this value,
  565. but only downward.
  566. </description>
  567. </property>
  568. <property>
  569. <name>mapreduce.job.ubertask.maxreduces</name>
  570. <value>1</value>
  571. <description>Threshold for number of reduces, beyond which job is considered
  572. too big for the ubertasking optimization. CURRENTLY THE CODE CANNOT SUPPORT
  573. MORE THAN ONE REDUCE and will ignore larger values. (Zero is a valid max,
  574. however.) Users may override this value, but only downward.
  575. </description>
  576. </property>
  577. <property>
  578. <name>mapreduce.job.ubertask.maxbytes</name>
  579. <value></value>
  580. <description>Threshold for number of input bytes, beyond which job is
  581. considered too big for the ubertasking optimization. If no value is
  582. specified, dfs.block.size is used as a default. Be sure to specify a
  583. default value in mapred-site.xml if the underlying filesystem is not HDFS.
  584. Users may override this value, but only downward.
  585. </description>
  586. </property>
  587. <property>
  588. <name>mapreduce.input.fileinputformat.split.minsize</name>
  589. <value>0</value>
  590. <description>The minimum size chunk that map input should be split
  591. into. Note that some file formats may have minimum split sizes that
  592. take priority over this setting.</description>
  593. </property>
  594. <property>
  595. <name>mapreduce.jobtracker.maxtasks.perjob</name>
  596. <value>-1</value>
  597. <description>The maximum number of tasks for a single job.
  598. A value of -1 indicates that there is no maximum. </description>
  599. </property>
  600. <property>
  601. <name>mapreduce.client.submit.file.replication</name>
  602. <value>10</value>
  603. <description>The replication level for submitted job files. This
  604. should be around the square root of the number of nodes.
  605. </description>
  606. </property>
  607. <property>
  608. <name>mapreduce.tasktracker.dns.interface</name>
  609. <value>default</value>
  610. <description>The name of the Network Interface from which a task
  611. tracker should report its IP address.
  612. </description>
  613. </property>
  614. <property>
  615. <name>mapreduce.tasktracker.dns.nameserver</name>
  616. <value>default</value>
  617. <description>The host name or IP address of the name server (DNS)
  618. which a TaskTracker should use to determine the host name used by
  619. the JobTracker for communication and display purposes.
  620. </description>
  621. </property>
  622. <property>
  623. <name>mapreduce.tasktracker.http.threads</name>
  624. <value>40</value>
  625. <description>The number of worker threads that for the http server. This is
  626. used for map output fetching
  627. </description>
  628. </property>
  629. <property>
  630. <name>mapreduce.tasktracker.http.address</name>
  631. <value>0.0.0.0:50060</value>
  632. <description>
  633. The task tracker http server address and port.
  634. If the port is 0 then the server will start on a free port.
  635. </description>
  636. </property>
  637. <property>
  638. <name>mapreduce.task.files.preserve.failedtasks</name>
  639. <value>false</value>
  640. <description>Should the files for failed tasks be kept. This should only be
  641. used on jobs that are failing, because the storage is never
  642. reclaimed. It also prevents the map outputs from being erased
  643. from the reduce directory as they are consumed.</description>
  644. </property>
  645. <!--
  646. <property>
  647. <name>mapreduce.task.files.preserve.filepattern</name>
  648. <value>.*_m_123456_0</value>
  649. <description>Keep all files from tasks whose task names match the given
  650. regular expression. Defaults to none.</description>
  651. </property>
  652. -->
  653. <property>
  654. <name>mapreduce.output.fileoutputformat.compress</name>
  655. <value>false</value>
  656. <description>Should the job outputs be compressed?
  657. </description>
  658. </property>
  659. <property>
  660. <name>mapreduce.output.fileoutputformat.compress.type</name>
  661. <value>RECORD</value>
  662. <description>If the job outputs are to compressed as SequenceFiles, how should
  663. they be compressed? Should be one of NONE, RECORD or BLOCK.
  664. </description>
  665. </property>
  666. <property>
  667. <name>mapreduce.output.fileoutputformat.compress.codec</name>
  668. <value>org.apache.hadoop.io.compress.DefaultCodec</value>
  669. <description>If the job outputs are compressed, how should they be compressed?
  670. </description>
  671. </property>
  672. <property>
  673. <name>mapreduce.map.output.compress</name>
  674. <value>false</value>
  675. <description>Should the outputs of the maps be compressed before being
  676. sent across the network. Uses SequenceFile compression.
  677. </description>
  678. </property>
  679. <property>
  680. <name>mapreduce.map.output.compress.codec</name>
  681. <value>org.apache.hadoop.io.compress.DefaultCodec</value>
  682. <description>If the map outputs are compressed, how should they be
  683. compressed?
  684. </description>
  685. </property>
  686. <property>
  687. <name>map.sort.class</name>
  688. <value>org.apache.hadoop.util.QuickSort</value>
  689. <description>The default sort class for sorting keys.
  690. </description>
  691. </property>
  692. <property>
  693. <name>mapreduce.task.userlog.limit.kb</name>
  694. <value>0</value>
  695. <description>The maximum size of user-logs of each task in KB. 0 disables the cap.
  696. </description>
  697. </property>
  698. <property>
  699. <name>mapreduce.job.userlog.retain.hours</name>
  700. <value>24</value>
  701. <description>The maximum time, in hours, for which the user-logs are to be
  702. retained after the job completion.
  703. </description>
  704. </property>
  705. <property>
  706. <name>mapreduce.jobtracker.hosts.filename</name>
  707. <value></value>
  708. <description>Names a file that contains the list of nodes that may
  709. connect to the jobtracker. If the value is empty, all hosts are
  710. permitted.</description>
  711. </property>
  712. <property>
  713. <name>mapreduce.jobtracker.hosts.exclude.filename</name>
  714. <value></value>
  715. <description>Names a file that contains the list of hosts that
  716. should be excluded by the jobtracker. If the value is empty, no
  717. hosts are excluded.</description>
  718. </property>
  719. <property>
  720. <name>mapreduce.jobtracker.heartbeats.in.second</name>
  721. <value>100</value>
  722. <description>Expert: Approximate number of heart-beats that could arrive
  723. at JobTracker in a second. Assuming each RPC can be processed
  724. in 10msec, the default value is made 100 RPCs in a second.
  725. </description>
  726. </property>
  727. <property>
  728. <name>mapreduce.jobtracker.tasktracker.maxblacklists</name>
  729. <value>4</value>
  730. <description>The number of blacklists for a taskTracker by various jobs
  731. after which the task tracker could be blacklisted across
  732. all jobs. The tracker will be given a tasks later
  733. (after a day). The tracker will become a healthy
  734. tracker after a restart.
  735. </description>
  736. </property>
  737. <property>
  738. <name>mapreduce.job.maxtaskfailures.per.tracker</name>
  739. <value>3</value>
  740. <description>The number of task-failures on a tasktracker of a given job
  741. after which new tasks of that job aren't assigned to it. It
  742. MUST be less than mapreduce.map.maxattempts and
  743. mapreduce.reduce.maxattempts otherwise the failed task will
  744. never be tried on a different node.
  745. </description>
  746. </property>
  747. <property>
  748. <name>mapreduce.client.output.filter</name>
  749. <value>FAILED</value>
  750. <description>The filter for controlling the output of the task's userlogs sent
  751. to the console of the JobClient.
  752. The permissible options are: NONE, KILLED, FAILED, SUCCEEDED and
  753. ALL.
  754. </description>
  755. </property>
  756. <property>
  757. <name>mapreduce.client.completion.pollinterval</name>
  758. <value>5000</value>
  759. <description>The interval (in milliseconds) between which the JobClient
  760. polls the JobTracker for updates about job status. You may want to set this
  761. to a lower value to make tests run faster on a single node system. Adjusting
  762. this value in production may lead to unwanted client-server traffic.
  763. </description>
  764. </property>
  765. <property>
  766. <name>mapreduce.client.progressmonitor.pollinterval</name>
  767. <value>1000</value>
  768. <description>The interval (in milliseconds) between which the JobClient
  769. reports status to the console and checks for job completion. You may want to set this
  770. to a lower value to make tests run faster on a single node system. Adjusting
  771. this value in production may lead to unwanted client-server traffic.
  772. </description>
  773. </property>
  774. <property>
  775. <name>mapreduce.jobtracker.persist.jobstatus.active</name>
  776. <value>true</value>
  777. <description>Indicates if persistency of job status information is
  778. active or not.
  779. </description>
  780. </property>
  781. <property>
  782. <name>mapreduce.jobtracker.persist.jobstatus.hours</name>
  783. <value>1</value>
  784. <description>The number of hours job status information is persisted in DFS.
  785. The job status information will be available after it drops of the memory
  786. queue and between jobtracker restarts. With a zero value the job status
  787. information is not persisted at all in DFS.
  788. </description>
  789. </property>
  790. <property>
  791. <name>mapreduce.jobtracker.persist.jobstatus.dir</name>
  792. <value>/jobtracker/jobsInfo</value>
  793. <description>The directory where the job status information is persisted
  794. in a file system to be available after it drops of the memory queue and
  795. between jobtracker restarts.
  796. </description>
  797. </property>
  798. <property>
  799. <name>mapreduce.task.profile</name>
  800. <value>false</value>
  801. <description>To set whether the system should collect profiler
  802. information for some of the tasks in this job? The information is stored
  803. in the user log directory. The value is "true" if task profiling
  804. is enabled.</description>
  805. </property>
  806. <property>
  807. <name>mapreduce.task.profile.maps</name>
  808. <value>0-2</value>
  809. <description> To set the ranges of map tasks to profile.
  810. mapreduce.task.profile has to be set to true for the value to be accounted.
  811. </description>
  812. </property>
  813. <property>
  814. <name>mapreduce.task.profile.reduces</name>
  815. <value>0-2</value>
  816. <description> To set the ranges of reduce tasks to profile.
  817. mapreduce.task.profile has to be set to true for the value to be accounted.
  818. </description>
  819. </property>
  820. <property>
  821. <name>mapreduce.task.skip.start.attempts</name>
  822. <value>2</value>
  823. <description> The number of Task attempts AFTER which skip mode
  824. will be kicked off. When skip mode is kicked off, the
  825. tasks reports the range of records which it will process
  826. next, to the TaskTracker. So that on failures, TT knows which
  827. ones are possibly the bad records. On further executions,
  828. those are skipped.
  829. </description>
  830. </property>
  831. <property>
  832. <name>mapreduce.map.skip.proc.count.autoincr</name>
  833. <value>true</value>
  834. <description> The flag which if set to true,
  835. SkipBadRecords.COUNTER_MAP_PROCESSED_RECORDS is incremented
  836. by MapRunner after invoking the map function. This value must be set to
  837. false for applications which process the records asynchronously
  838. or buffer the input records. For example streaming.
  839. In such cases applications should increment this counter on their own.
  840. </description>
  841. </property>
  842. <property>
  843. <name>mapreduce.reduce.skip.proc.count.autoincr</name>
  844. <value>true</value>
  845. <description> The flag which if set to true,
  846. SkipBadRecords.COUNTER_REDUCE_PROCESSED_GROUPS is incremented
  847. by framework after invoking the reduce function. This value must be set to
  848. false for applications which process the records asynchronously
  849. or buffer the input records. For example streaming.
  850. In such cases applications should increment this counter on their own.
  851. </description>
  852. </property>
  853. <property>
  854. <name>mapreduce.job.skip.outdir</name>
  855. <value></value>
  856. <description> If no value is specified here, the skipped records are
  857. written to the output directory at _logs/skip.
  858. User can stop writing skipped records by giving the value "none".
  859. </description>
  860. </property>
  861. <property>
  862. <name>mapreduce.map.skip.maxrecords</name>
  863. <value>0</value>
  864. <description> The number of acceptable skip records surrounding the bad
  865. record PER bad record in mapper. The number includes the bad record as well.
  866. To turn the feature of detection/skipping of bad records off, set the
  867. value to 0.
  868. The framework tries to narrow down the skipped range by retrying
  869. until this threshold is met OR all attempts get exhausted for this task.
  870. Set the value to Long.MAX_VALUE to indicate that framework need not try to
  871. narrow down. Whatever records(depends on application) get skipped are
  872. acceptable.
  873. </description>
  874. </property>
  875. <property>
  876. <name>mapreduce.reduce.skip.maxgroups</name>
  877. <value>0</value>
  878. <description> The number of acceptable skip groups surrounding the bad
  879. group PER bad group in reducer. The number includes the bad group as well.
  880. To turn the feature of detection/skipping of bad groups off, set the
  881. value to 0.
  882. The framework tries to narrow down the skipped range by retrying
  883. until this threshold is met OR all attempts get exhausted for this task.
  884. Set the value to Long.MAX_VALUE to indicate that framework need not try to
  885. narrow down. Whatever groups(depends on application) get skipped are
  886. acceptable.
  887. </description>
  888. </property>
  889. <property>
  890. <name>mapreduce.ifile.readahead</name>
  891. <value>true</value>
  892. <description>Configuration key to enable/disable IFile readahead.
  893. </description>
  894. </property>
  895. <property>
  896. <name>mapreduce.ifile.readahead.bytes</name>
  897. <value>4194304</value>
  898. <description>Configuration key to set the IFile readahead length in bytes.
  899. </description>
  900. </property>
  901. <!-- Proxy Configuration -->
  902. <property>
  903. <name>mapreduce.jobtracker.taskcache.levels</name>
  904. <value>2</value>
  905. <description> This is the max level of the task cache. For example, if
  906. the level is 2, the tasks cached are at the host level and at the rack
  907. level.
  908. </description>
  909. </property>
  910. <property>
  911. <name>mapreduce.job.queuename</name>
  912. <value>default</value>
  913. <description> Queue to which a job is submitted. This must match one of the
  914. queues defined in mapred-queues.xml for the system. Also, the ACL setup
  915. for the queue must allow the current user to submit a job to the queue.
  916. Before specifying a queue, ensure that the system is configured with
  917. the queue, and access is allowed for submitting jobs to the queue.
  918. </description>
  919. </property>
  920. <property>
  921. <name>mapreduce.cluster.acls.enabled</name>
  922. <value>false</value>
  923. <description> Specifies whether ACLs should be checked
  924. for authorization of users for doing various queue and job level operations.
  925. ACLs are disabled by default. If enabled, access control checks are made by
  926. JobTracker and TaskTracker when requests are made by users for queue
  927. operations like submit job to a queue and kill a job in the queue and job
  928. operations like viewing the job-details (See mapreduce.job.acl-view-job)
  929. or for modifying the job (See mapreduce.job.acl-modify-job) using
  930. Map/Reduce APIs, RPCs or via the console and web user interfaces.
  931. For enabling this flag(mapreduce.cluster.acls.enabled), this is to be set
  932. to true in mapred-site.xml on JobTracker node and on all TaskTracker nodes.
  933. </description>
  934. </property>
  935. <property>
  936. <name>mapreduce.job.acl-modify-job</name>
  937. <value> </value>
  938. <description> Job specific access-control list for 'modifying' the job. It
  939. is only used if authorization is enabled in Map/Reduce by setting the
  940. configuration property mapreduce.cluster.acls.enabled to true.
  941. This specifies the list of users and/or groups who can do modification
  942. operations on the job. For specifying a list of users and groups the
  943. format to use is "user1,user2 group1,group". If set to '*', it allows all
  944. users/groups to modify this job. If set to ' '(i.e. space), it allows
  945. none. This configuration is used to guard all the modifications with respect
  946. to this job and takes care of all the following operations:
  947. o killing this job
  948. o killing a task of this job, failing a task of this job
  949. o setting the priority of this job
  950. Each of these operations are also protected by the per-queue level ACL
  951. "acl-administer-jobs" configured via mapred-queues.xml. So a caller should
  952. have the authorization to satisfy either the queue-level ACL or the
  953. job-level ACL.
  954. Irrespective of this ACL configuration, (a) job-owner, (b) the user who
  955. started the cluster, (c) members of an admin configured supergroup
  956. configured via mapreduce.cluster.permissions.supergroup and (d) queue
  957. administrators of the queue to which this job was submitted to configured
  958. via acl-administer-jobs for the specific queue in mapred-queues.xml can
  959. do all the modification operations on a job.
  960. By default, nobody else besides job-owner, the user who started the cluster,
  961. members of supergroup and queue administrators can perform modification
  962. operations on a job.
  963. </description>
  964. </property>
  965. <property>
  966. <name>mapreduce.job.acl-view-job</name>
  967. <value> </value>
  968. <description> Job specific access-control list for 'viewing' the job. It is
  969. only used if authorization is enabled in Map/Reduce by setting the
  970. configuration property mapreduce.cluster.acls.enabled to true.
  971. This specifies the list of users and/or groups who can view private details
  972. about the job. For specifying a list of users and groups the
  973. format to use is "user1,user2 group1,group". If set to '*', it allows all
  974. users/groups to modify this job. If set to ' '(i.e. space), it allows
  975. none. This configuration is used to guard some of the job-views and at
  976. present only protects APIs that can return possibly sensitive information
  977. of the job-owner like
  978. o job-level counters
  979. o task-level counters
  980. o tasks' diagnostic information
  981. o task-logs displayed on the TaskTracker web-UI and
  982. o job.xml showed by the JobTracker's web-UI
  983. Every other piece of information of jobs is still accessible by any other
  984. user, for e.g., JobStatus, JobProfile, list of jobs in the queue, etc.
  985. Irrespective of this ACL configuration, (a) job-owner, (b) the user who
  986. started the cluster, (c) members of an admin configured supergroup
  987. configured via mapreduce.cluster.permissions.supergroup and (d) queue
  988. administrators of the queue to which this job was submitted to configured
  989. via acl-administer-jobs for the specific queue in mapred-queues.xml can
  990. do all the view operations on a job.
  991. By default, nobody else besides job-owner, the user who started the
  992. cluster, memebers of supergroup and queue administrators can perform
  993. view operations on a job.
  994. </description>
  995. </property>
  996. <property>
  997. <name>mapreduce.tasktracker.indexcache.mb</name>
  998. <value>10</value>
  999. <description> The maximum memory that a task tracker allows for the
  1000. index cache that is used when serving map outputs to reducers.
  1001. </description>
  1002. </property>
  1003. <property>
  1004. <name>mapreduce.job.token.tracking.ids.enabled</name>
  1005. <value>false</value>
  1006. <description>Whether to write tracking ids of tokens to
  1007. job-conf. When true, the configuration property
  1008. "mapreduce.job.token.tracking.ids" is set to the token-tracking-ids of
  1009. the job</description>
  1010. </property>
  1011. <property>
  1012. <name>mapreduce.job.token.tracking.ids</name>
  1013. <value></value>
  1014. <description>When mapreduce.job.token.tracking.ids.enabled is
  1015. set to true, this is set by the framework to the
  1016. token-tracking-ids used by the job.</description>
  1017. </property>
  1018. <property>
  1019. <name>mapreduce.task.merge.progress.records</name>
  1020. <value>10000</value>
  1021. <description> The number of records to process during merge before
  1022. sending a progress notification to the TaskTracker.
  1023. </description>
  1024. </property>
  1025. <property>
  1026. <name>mapreduce.job.reduce.slowstart.completedmaps</name>
  1027. <value>0.05</value>
  1028. <description>Fraction of the number of maps in the job which should be
  1029. complete before reduces are scheduled for the job.
  1030. </description>
  1031. </property>
  1032. <property>
  1033. <name>mapreduce.job.complete.cancel.delegation.tokens</name>
  1034. <value>true</value>
  1035. <description> if false - do not unregister/cancel delegation tokens from
  1036. renewal, because same tokens may be used by spawned jobs
  1037. </description>
  1038. </property>
  1039. <property>
  1040. <name>mapreduce.tasktracker.taskcontroller</name>
  1041. <value>org.apache.hadoop.mapred.DefaultTaskController</value>
  1042. <description>TaskController which is used to launch and manage task execution
  1043. </description>
  1044. </property>
  1045. <property>
  1046. <name>mapreduce.tasktracker.group</name>
  1047. <value></value>
  1048. <description>Expert: Group to which TaskTracker belongs. If
  1049. LinuxTaskController is configured via mapreduce.tasktracker.taskcontroller,
  1050. the group owner of the task-controller binary should be same as this group.
  1051. </description>
  1052. </property>
  1053. <property>
  1054. <name>mapreduce.shuffle.port</name>
  1055. <value>13562</value>
  1056. <description>Default port that the ShuffleHandler will run on. ShuffleHandler
  1057. is a service run at the NodeManager to facilitate transfers of intermediate
  1058. Map outputs to requesting Reducers.
  1059. </description>
  1060. </property>
  1061. <property>
  1062. <name>mapreduce.job.reduce.shuffle.consumer.plugin.class</name>
  1063. <value>org.apache.hadoop.mapreduce.task.reduce.Shuffle</value>
  1064. <description>
  1065. Name of the class whose instance will be used
  1066. to send shuffle requests by reducetasks of this job.
  1067. The class must be an instance of org.apache.hadoop.mapred.ShuffleConsumerPlugin.
  1068. </description>
  1069. </property>
  1070. <!-- Node health script variables -->
  1071. <property>
  1072. <name>mapreduce.tasktracker.healthchecker.script.path</name>
  1073. <value></value>
  1074. <description>Absolute path to the script which is
  1075. periodicallyrun by the node health monitoring service to determine if
  1076. the node is healthy or not. If the value of this key is empty or the
  1077. file does not exist in the location configured here, the node health
  1078. monitoring service is not started.</description>
  1079. </property>
  1080. <property>
  1081. <name>mapreduce.tasktracker.healthchecker.interval</name>
  1082. <value>60000</value>
  1083. <description>Frequency of the node health script to be run,
  1084. in milliseconds</description>
  1085. </property>
  1086. <property>
  1087. <name>mapreduce.tasktracker.healthchecker.script.timeout</name>
  1088. <value>600000</value>
  1089. <description>Time after node health script should be killed if
  1090. unresponsive and considered that the script has failed.</description>
  1091. </property>
  1092. <property>
  1093. <name>mapreduce.tasktracker.healthchecker.script.args</name>
  1094. <value></value>
  1095. <description>List of arguments which are to be passed to
  1096. node health script when it is being launched comma seperated.
  1097. </description>
  1098. </property>
  1099. <!-- end of node health script variables -->
  1100. <!-- MR YARN Application properties -->
  1101. <property>
  1102. <name>mapreduce.job.counters.limit</name>
  1103. <value>120</value>
  1104. <description>Limit on the number of user counters allowed per job.
  1105. </description>
  1106. </property>
  1107. <property>
  1108. <name>mapreduce.framework.name</name>
  1109. <value>local</value>
  1110. <description>The runtime framework for executing MapReduce jobs.
  1111. Can be one of local, classic or yarn.
  1112. </description>
  1113. </property>
  1114. <property>
  1115. <name>yarn.app.mapreduce.am.staging-dir</name>
  1116. <value>/tmp/hadoop-yarn/staging</value>
  1117. <description>The staging dir used while submitting jobs.
  1118. </description>
  1119. </property>
  1120. <property>
  1121. <name>mapreduce.am.max-attempts</name>
  1122. <value>2</value>
  1123. <description>The maximum number of application attempts. It is a
  1124. application-specific setting. It should not be larger than the global number
  1125. set by resourcemanager. Otherwise, it will be override. The default number is
  1126. set to 2, to allow at least one retry for AM.</description>
  1127. </property>
  1128. <!-- Job Notification Configuration -->
  1129. <property>
  1130. <name>mapreduce.job.end-notification.url</name>
  1131. <!--<value>http://localhost:8080/jobstatus.php?jobId=$jobId&amp;jobStatus=$jobStatus</value>-->
  1132. <description>Indicates url which will be called on completion of job to inform
  1133. end status of job.
  1134. User can give at most 2 variables with URI : $jobId and $jobStatus.
  1135. If they are present in URI, then they will be replaced by their
  1136. respective values.
  1137. </description>
  1138. </property>
  1139. <property>
  1140. <name>mapreduce.job.end-notification.retry.attempts</name>
  1141. <value>0</value>
  1142. <description>The number of times the submitter of the job wants to retry job
  1143. end notification if it fails. This is capped by
  1144. mapreduce.job.end-notification.max.attempts</description>
  1145. </property>
  1146. <property>
  1147. <name>mapreduce.job.end-notification.retry.interval</name>
  1148. <value>1000</value>
  1149. <description>The number of milliseconds the submitter of the job wants to
  1150. wait before job end notification is retried if it fails. This is capped by
  1151. mapreduce.job.end-notification.max.retry.interval</description>
  1152. </property>
  1153. <property>
  1154. <name>mapreduce.job.end-notification.max.attempts</name>
  1155. <value>5</value>
  1156. <final>true</final>
  1157. <description>The maximum number of times a URL will be read for providing job
  1158. end notification. Cluster administrators can set this to limit how long
  1159. after end of a job, the Application Master waits before exiting. Must be
  1160. marked as final to prevent users from overriding this.
  1161. </description>
  1162. </property>
  1163. <property>
  1164. <name>mapreduce.job.end-notification.max.retry.interval</name>
  1165. <value>5000</value>
  1166. <final>true</final>
  1167. <description>The maximum amount of time (in milliseconds) to wait before
  1168. retrying job end notification. Cluster administrators can set this to
  1169. limit how long the Application Master waits before exiting. Must be marked
  1170. as final to prevent users from overriding this.</description>
  1171. </property>
  1172. <property>
  1173. <name>yarn.app.mapreduce.am.env</name>
  1174. <value></value>
  1175. <description>User added environment variables for the MR App Master
  1176. processes. Example :
  1177. 1) A=foo This will set the env variable A to foo
  1178. 2) B=$B:c This is inherit tasktracker's B env variable.
  1179. </description>
  1180. </property>
  1181. <property>
  1182. <name>yarn.app.mapreduce.am.admin.user.env</name>
  1183. <value></value>
  1184. <description> Environment variables for the MR App Master
  1185. processes for admin purposes. These values are set first and can be
  1186. overridden by the user env (yarn.app.mapreduce.am.env) Example :
  1187. 1) A=foo This will set the env variable A to foo
  1188. 2) B=$B:c This is inherit app master's B env variable.
  1189. </description>
  1190. </property>
  1191. <property>
  1192. <name>yarn.app.mapreduce.am.command-opts</name>
  1193. <value>-Xmx1024m</value>
  1194. <description>Java opts for the MR App Master processes.
  1195. The following symbol, if present, will be interpolated: @taskid@ is replaced
  1196. by current TaskID. Any other occurrences of '@' will go unchanged.
  1197. For example, to enable verbose gc logging to a file named for the taskid in
  1198. /tmp and to set the heap maximum to be a gigabyte, pass a 'value' of:
  1199. -Xmx1024m -verbose:gc -Xloggc:/tmp/@taskid@.gc
  1200. Usage of -Djava.library.path can cause programs to no longer function if
  1201. hadoop native libraries are used. These values should instead be set as part
  1202. of LD_LIBRARY_PATH in the map / reduce JVM env using the mapreduce.map.env and
  1203. mapreduce.reduce.env config settings.
  1204. </description>
  1205. </property>
  1206. <property>
  1207. <name>yarn.app.mapreduce.am.admin-command-opts</name>
  1208. <value></value>
  1209. <description>Java opts for the MR App Master processes for admin purposes.
  1210. It will appears before the opts set by yarn.app.mapreduce.am.command-opts and
  1211. thus its options can be overridden user.
  1212. Usage of -Djava.library.path can cause programs to no longer function if
  1213. hadoop native libraries are used. These values should instead be set as part
  1214. of LD_LIBRARY_PATH in the map / reduce JVM env using the mapreduce.map.env and
  1215. mapreduce.reduce.env config settings.
  1216. </description>
  1217. </property>
  1218. <property>
  1219. <name>yarn.app.mapreduce.am.job.task.listener.thread-count</name>
  1220. <value>30</value>
  1221. <description>The number of threads used to handle RPC calls in the
  1222. MR AppMaster from remote tasks</description>
  1223. </property>
  1224. <property>
  1225. <name>yarn.app.mapreduce.am.job.client.port-range</name>
  1226. <value></value>
  1227. <description>Range of ports that the MapReduce AM can use when binding.
  1228. Leave blank if you want all possible ports.
  1229. For example 50000-50050,50100-50200</description>
  1230. </property>
  1231. <property>
  1232. <name>yarn.app.mapreduce.am.job.committer.cancel-timeout</name>
  1233. <value>60000</value>
  1234. <description>The amount of time in milliseconds to wait for the output
  1235. committer to cancel an operation if the job is killed</description>
  1236. </property>
  1237. <property>
  1238. <name>yarn.app.mapreduce.am.job.committer.commit-window</name>
  1239. <value>10000</value>
  1240. <description>Defines a time window in milliseconds for output commit
  1241. operations. If contact with the RM has occurred within this window then
  1242. commits are allowed, otherwise the AM will not allow output commits until
  1243. contact with the RM has been re-established.</description>
  1244. </property>
  1245. <property>
  1246. <name>yarn.app.mapreduce.am.scheduler.heartbeat.interval-ms</name>
  1247. <value>1000</value>
  1248. <description>The interval in ms at which the MR AppMaster should send
  1249. heartbeats to the ResourceManager</description>
  1250. </property>
  1251. <property>
  1252. <name>yarn.app.mapreduce.client-am.ipc.max-retries</name>
  1253. <value>1</value>
  1254. <description>The number of client retries to the AM - before reconnecting
  1255. to the RM to fetch Application Status.</description>
  1256. </property>
  1257. <property>
  1258. <name>yarn.app.mapreduce.client.max-retries</name>
  1259. <value>3</value>
  1260. <description>The number of client retries to the RM/HS/AM before
  1261. throwing exception. This is a layer above the ipc.</description>
  1262. </property>
  1263. <property>
  1264. <name>yarn.app.mapreduce.am.resource.mb</name>
  1265. <value>1536</value>
  1266. <description>The amount of memory the MR AppMaster needs.</description>
  1267. </property>
  1268. <property>
  1269. <name>yarn.app.mapreduce.am.resource.cpu-vcores</name>
  1270. <value>1</value>
  1271. <description>
  1272. The number of virtual CPU cores the MR AppMaster needs.
  1273. </description>
  1274. </property>
  1275. <property>
  1276. <description>CLASSPATH for MR applications. A comma-separated list
  1277. of CLASSPATH entries</description>
  1278. <name>mapreduce.application.classpath</name>
  1279. <value>$HADOOP_MAPRED_HOME/share/hadoop/mapreduce/*,$HADOOP_MAPRED_HOME/share/hadoop/mapreduce/lib/*</value>
  1280. </property>
  1281. <property>
  1282. <name>mapreduce.job.classloader</name>
  1283. <value>false</value>
  1284. <description>Whether to use a separate (isolated) classloader for
  1285. user classes in the task JVM.</description>
  1286. </property>
  1287. <property>
  1288. <name>mapreduce.job.classloader.system.classes</name>
  1289. <value>java.,javax.,org.apache.commons.logging.,org.apache.log4j.,org.apache.hadoop.</value>
  1290. <description>A comma-separated list of classes that should be loaded from the
  1291. system classpath, not the user-supplied JARs, when mapreduce.job.classloader
  1292. is enabled. Names ending in '.' (period) are treated as package names,
  1293. and names starting with a '-' are treated as negative matches.
  1294. </description>
  1295. </property>
  1296. <!-- jobhistory properties -->
  1297. <property>
  1298. <name>mapreduce.jobhistory.address</name>
  1299. <value>0.0.0.0:10020</value>
  1300. <description>MapReduce JobHistory Server IPC host:port</description>
  1301. </property>
  1302. <property>
  1303. <name>mapreduce.jobhistory.webapp.address</name>
  1304. <value>0.0.0.0:19888</value>
  1305. <description>MapReduce JobHistory Server Web UI host:port</description>
  1306. </property>
  1307. <property>
  1308. <name>mapreduce.jobhistory.keytab</name>
  1309. <description>
  1310. Location of the kerberos keytab file for the MapReduce
  1311. JobHistory Server.
  1312. </description>
  1313. <value>/etc/security/keytab/jhs.service.keytab</value>
  1314. </property>
  1315. <property>
  1316. <name>mapreduce.jobhistory.principal</name>
  1317. <description>
  1318. Kerberos principal name for the MapReduce JobHistory Server.
  1319. </description>
  1320. <value>jhs/_HOST@REALM.TLD</value>
  1321. </property>
  1322. <property>
  1323. <name>mapreduce.jobhistory.intermediate-done-dir</name>
  1324. <value>${yarn.app.mapreduce.am.staging-dir}/history/done_intermediate</value>
  1325. <description></description>
  1326. </property>
  1327. <property>
  1328. <name>mapreduce.jobhistory.done-dir</name>
  1329. <value>${yarn.app.mapreduce.am.staging-dir}/history/done</value>
  1330. <description></description>
  1331. </property>
  1332. <property>
  1333. <name>mapreduce.jobhistory.cleaner.enable</name>
  1334. <value>true</value>
  1335. <description></description>
  1336. </property>
  1337. <property>
  1338. <name>mapreduce.jobhistory.cleaner.interval-ms</name>
  1339. <value>86400000</value>
  1340. <description> How often the job history cleaner checks for files to delete,
  1341. in milliseconds. Defaults to 86400000 (one day). Files are only deleted if
  1342. they are older than mapreduce.jobhistory.max-age-ms.
  1343. </description>
  1344. </property>
  1345. <property>
  1346. <name>mapreduce.jobhistory.max-age-ms</name>
  1347. <value>604800000</value>
  1348. <description> Job history files older than this many milliseconds will
  1349. be deleted when the history cleaner runs. Defaults to 604800000 (1 week).
  1350. </description>
  1351. </property>
  1352. <property>
  1353. <name>mapreduce.jobhistory.client.thread-count</name>
  1354. <value>10</value>
  1355. <description>The number of threads to handle client API requests</description>
  1356. </property>
  1357. <property>
  1358. <name>mapreduce.jobhistory.datestring.cache.size</name>
  1359. <value>200000</value>
  1360. <description>Size of the date string cache. Effects the number of directories
  1361. which will be scanned to find a job.</description>
  1362. </property>
  1363. <property>
  1364. <name>mapreduce.jobhistory.joblist.cache.size</name>
  1365. <value>20000</value>
  1366. <description>Size of the job list cache</description>
  1367. </property>
  1368. <property>
  1369. <name>mapreduce.jobhistory.loadedjobs.cache.size</name>
  1370. <value>5</value>
  1371. <description>Size of the loaded job cache</description>
  1372. </property>
  1373. <property>
  1374. <name>mapreduce.jobhistory.move.interval-ms</name>
  1375. <value>180000</value>
  1376. <description>Scan for history files to more from intermediate done dir to done
  1377. dir at this frequency.
  1378. </description>
  1379. </property>
  1380. <property>
  1381. <name>mapreduce.jobhistory.move.thread-count</name>
  1382. <value>3</value>
  1383. <description>The number of threads used to move files.</description>
  1384. </property>
  1385. <property>
  1386. <name>mapreduce.jobhistory.store.class</name>
  1387. <value></value>
  1388. <description>The HistoryStorage class to use to cache history data.</description>
  1389. </property>
  1390. <property>
  1391. <name>mapreduce.jobhistory.minicluster.fixed.ports</name>
  1392. <value>false</value>
  1393. <description>Whether to use fixed ports with the minicluster</description>
  1394. </property>
  1395. <property>
  1396. <name>mapreduce.jobhistory.admin.address</name>
  1397. <value>0.0.0.0:10033</value>
  1398. <description>The address of the History server admin interface.</description>
  1399. </property>
  1400. <property>
  1401. <name>mapreduce.jobhistory.admin.acl</name>
  1402. <value>*</value>
  1403. <description>ACL of who can be admin of the History server.</description>
  1404. </property>
  1405. </configuration>