1 <?xml version="1.0"?> 2 <?xml-stylesheet type="text/xsl" href="configuration.xsl"?> 3 <!-- 4 /** 5 * Copyright 2009 The Apache Software Foundation 6 * 7 * Licensed to the Apache Software Foundation (ASF) under one 8 * or more contributor license agreements. See the NOTICE file 9 * distributed with this work for additional information 10 * regarding copyright ownership. The ASF licenses this file 11 * to you under the Apache License, Version 2.0 (the 12 * "License"); you may not use this file except in compliance 13 * with the License. You may obtain a copy of the License at 14 * 15 * http://www.apache.org/licenses/LICENSE-2.0 16 * 17 * Unless required by applicable law or agreed to in writing, software 18 * distributed under the License is distributed on an "AS IS" BASIS, 19 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 20 * See the License for the specific language governing permissions and 21 * limitations under the License. 22 */ 23 --> 24 <configuration> 25 <property> 26 <name>hbase.rootdir</name> 27 <value>file:///tmp/hbase-${user.name}/hbase</value> 28 <description>The directory shared by region servers and into 29 which HBase persists. The URL should be 'fully-qualified' 30 to include the filesystem scheme. For example, to specify the 31 HDFS directory '/hbase' where the HDFS instance's namenode is 32 running at namenode.example.org on port 9000, set this value to: 33 hdfs://namenode.example.org:9000/hbase. By default HBase writes 34 into /tmp. Change this configuration else all data will be lost 35 on machine restart. 36 </description> 37 </property> 38 <property> 39 <name>hbase.master.port</name> 40 <value>60000</value> 41 <description>The port the HBase Master should bind to.</description> 42 </property> 43 <property> 44 <name>hbase.cluster.distributed</name> 45 <value>false</value> 46 <description>The mode the cluster will be in. Possible values are 47 false for standalone mode and true for distributed mode. If 48 false, startup will run all HBase and ZooKeeper daemons together 49 in the one JVM. 50 </description> 51 </property> 52 <property> 53 <name>hbase.tmp.dir</name> 54 <value>/tmp/hbase-${user.name}</value> 55 <description>Temporary directory on the local filesystem. 56 Change this setting to point to a location more permanent 57 than '/tmp' (The '/tmp' directory is often cleared on 58 machine restart). 59 </description> 60 </property> 61 <property> 62 <name>hbase.master.info.port</name> 63 <value>60010</value> 64 <description>The port for the HBase Master web UI. 65 Set to -1 if you do not want a UI instance run. 66 </description> 67 </property> 68 <property> 69 <name>hbase.master.info.bindAddress</name> 70 <value>0.0.0.0</value> 71 <description>The bind address for the HBase Master web UI 72 </description> 73 </property> 74 <property> 75 <name>hbase.client.write.buffer</name> 76 <value>2097152</value> 77 <description>Default size of the HTable clien write buffer in bytes. 78 A bigger buffer takes more memory -- on both the client and server 79 side since server instantiates the passed write buffer to process 80 it -- but a larger buffer size reduces the number of RPCs made. 81 For an estimate of server-side memory-used, evaluate 82 hbase.client.write.buffer * hbase.regionserver.handler.count 83 </description> 84 </property> 85 <property> 86 <name>hbase.regionserver.port</name> 87 <value>60020</value> 88 <description>The port the HBase RegionServer binds to. 89 </description> 90 </property> 91 <property> 92 <name>hbase.regionserver.info.port</name> 93 <value>60030</value> 94 <description>The port for the HBase RegionServer web UI 95 Set to -1 if you do not want the RegionServer UI to run. 96 </description> 97 </property> 98 <property> 99 <name>hbase.regionserver.info.port.auto</name> 100 <value>false</value> 101 <description>Whether or not the Master or RegionServer 102 UI should search for a port to bind to. Enables automatic port 103 search if hbase.regionserver.info.port is already in use. 104 Useful for testing, turned off by default. 105 </description> 106 </property> 107 <property> 108 <name>hbase.regionserver.info.bindAddress</name> 109 <value>0.0.0.0</value> 110 <description>The address for the HBase RegionServer web UI 111 </description> 112 </property> 113 <property> 114 <name>hbase.regionserver.class</name> 115 <value>org.apache.hadoop.hbase.ipc.HRegionInterface</value> 116 <description>The RegionServer interface to use. 117 Used by the client opening proxy to remote region server. 118 </description> 119 </property> 120 <property> 121 <name>hbase.client.pause</name> 122 <value>1000</value> 123 <description>General client pause value. Used mostly as value to wait 124 before running a retry of a failed get, region lookup, etc.</description> 125 </property> 126 <property> 127 <name>hbase.client.retries.number</name> 128 <value>10</value> 129 <description>Maximum retries. Used as maximum for all retryable 130 operations such as fetching of the root region from root region 131 server, getting a cell's value, starting a row update, etc. 132 Default: 10. 133 </description> 134 </property> 135 <property> 136 <name>hbase.client.scanner.caching</name> 137 <value>1</value> 138 <description>Number of rows that will be fetched when calling next 139 on a scanner if it is not served from (local, client) memory. Higher 140 caching values will enable faster scanners but will eat up more memory 141 and some calls of next may take longer and longer times when the cache is empty. 142 Do not set this value such that the time between invocations is greater 143 than the scanner timeout; i.e. hbase.regionserver.lease.period 144 </description> 145 </property> 146 <property> 147 <name>hbase.client.keyvalue.maxsize</name> 148 <value>10485760</value> 149 <description>Specifies the combined maximum allowed size of a KeyValue 150 instance. This is to set an upper boundary for a single entry saved in a 151 storage file. Since they cannot be split it helps avoiding that a region 152 cannot be split any further because the data is too large. It seems wise 153 to set this to a fraction of the maximum region size. Setting it to zero 154 or less disables the check. 155 </description> 156 </property> 157 <property> 158 <name>hbase.regionserver.lease.period</name> 159 <value>60000</value> 160 <description>HRegion server lease period in milliseconds. Default is 161 60 seconds. Clients must report in within this period else they are 162 considered dead.</description> 163 </property> 164 <property> 165 <name>hbase.regionserver.handler.count</name> 166 <value>10</value> 167 <description>Count of RPC Server instances spun up on RegionServers 168 Same property is used by the Master for count of master handlers. 169 Default is 10. 170 </description> 171 </property> 172 <property> 173 <name>hbase.regionserver.msginterval</name> 174 <value>3000</value> 175 <description>Interval between messages from the RegionServer to Master 176 in milliseconds. 177 </description> 178 </property> 179 <property> 180 <name>hbase.regionserver.flushlogentries</name> 181 <value>1</value> 182 <description>Sync the HLog to HDFS when it has accumulated this many 183 entries. Default 1. Value is checked on every HLog.hflush 184 </description> 185 </property> 186 <property> 187 <name>hbase.regionserver.optionallogflushinterval</name> 188 <value>1000</value> 189 <description>Sync the HLog to the HDFS after this interval if it has not 190 accumulated enough entries to trigger a sync. Default 1 second. Units: 191 milliseconds. 192 </description> 193 </property> 194 <property> 195 <name>hbase.regionserver.regionSplitLimit</name> 196 <value>2147483647</value> 197 <description>Limit for the number of regions after which no more region 198 splitting should take place. This is not a hard limit for the number of 199 regions but acts as a guideline for the regionserver to stop splitting after 200 a certain limit. Default is set to MAX_INT; i.e. do not block splitting. 201 </description> 202 </property> 203 <property> 204 <name>hbase.regionserver.logroll.period</name> 205 <value>3600000</value> 206 <description>Period at which we will roll the commit log regardless 207 of how many edits it has.</description> 208 </property> 209 <property> 210 <name>hbase.regionserver.hlog.reader.impl</name> 211 <value>org.apache.hadoop.hbase.regionserver.wal.SequenceFileLogReader</value> 212 <description>The HLog file reader implementation.</description> 213 </property> 214 <property> 215 <name>hbase.regionserver.hlog.writer.impl</name> 216 <value>org.apache.hadoop.hbase.regionserver.wal.SequenceFileLogWriter</value> 217 <description>The HLog file writer implementation.</description> 218 </property> 219 <property> 220 <name>hbase.regionserver.thread.splitcompactcheckfrequency</name> 221 <value>20000</value> 222 <description>How often a region server runs the split/compaction check. 223 </description> 224 </property> 225 <property> 226 <name>hbase.regionserver.nbreservationblocks</name> 227 <value>4</value> 228 <description>The number of resevoir blocks of memory release on 229 OOME so we can cleanup properly before server shutdown. 230 </description> 231 </property> 232 <property> 233 <name>hbase.zookeeper.dns.interface</name> 234 <value>default</value> 235 <description>The name of the Network Interface from which a ZooKeeper server 236 should report its IP address. 237 </description> 238 </property> 239 <property> 240 <name>hbase.zookeeper.dns.nameserver</name> 241 <value>default</value> 242 <description>The host name or IP address of the name server (DNS) 243 which a ZooKeeper server should use to determine the host name used by the 244 master for communication and display purposes. 245 </description> 246 </property> 247 <property> 248 <name>hbase.regionserver.dns.interface</name> 249 <value>default</value> 250 <description>The name of the Network Interface from which a region server 251 should report its IP address. 252 </description> 253 </property> 254 <property> 255 <name>hbase.regionserver.dns.nameserver</name> 256 <value>default</value> 257 <description>The host name or IP address of the name server (DNS) 258 which a region server should use to determine the host name used by the 259 master for communication and display purposes. 260 </description> 261 </property> 262 <property> 263 <name>hbase.master.dns.interface</name> 264 <value>default</value> 265 <description>The name of the Network Interface from which a master 266 should report its IP address. 267 </description> 268 </property> 269 <property> 270 <name>hbase.master.dns.nameserver</name> 271 <value>default</value> 272 <description>The host name or IP address of the name server (DNS) 273 which a master should use to determine the host name used 274 for communication and display purposes. 275 </description> 276 </property> 277 <property> 278 <name>hbase.balancer.period 279 </name> 280 <value>300000</value> 281 <description>Period at which the region balancer runs in the Master. 282 </description> 283 </property> 284 <property> 285 <name>hbase.master.logcleaner.ttl</name> 286 <value>600000</value> 287 <description>Maximum time a HLog can stay in the .oldlogdir directory, 288 after which it will be cleaned by a Master thread. 289 </description> 290 </property> 291 <property> 292 <name>hbase.master.logcleaner.plugins</name> 293 <value>org.apache.hadoop.hbase.master.TimeToLiveLogCleaner</value> 294 <description>A comma-separated list of LogCleanerDelegate invoked by 295 the LogsCleaner service. These WAL/HLog cleaners are called in order, 296 so put the HLog cleaner that prunes the most HLog files in front. To 297 implement your own LogCleanerDelegate, just put it in HBase's classpath 298 and add the fully qualified class name here. Always add the above 299 default log cleaners in the list. 300 </description> 301 </property> 302 <property> 303 <name>hbase.regionserver.global.memstore.upperLimit</name> 304 <value>0.4</value> 305 <description>Maximum size of all memstores in a region server before new 306 updates are blocked and flushes are forced. Defaults to 40% of heap 307 </description> 308 </property> 309 <property> 310 <name>hbase.regionserver.global.memstore.lowerLimit</name> 311 <value>0.35</value> 312 <description>When memstores are being forced to flush to make room in 313 memory, keep flushing until we hit this mark. Defaults to 35% of heap. 314 This value equal to hbase.regionserver.global.memstore.upperLimit causes 315 the minimum possible flushing to occur when updates are blocked due to 316 memstore limiting. 317 </description> 318 </property> 319 <property> 320 <name>hbase.server.thread.wakefrequency</name> 321 <value>10000</value> 322 <description>Time to sleep in between searches for work (in milliseconds). 323 Used as sleep interval by service threads such as log roller. 324 </description> 325 </property> 326 <property> 327 <name>hbase.hregion.memstore.flush.size</name> 328 <value>67108864</value> 329 <description> 330 Memstore will be flushed to disk if size of the memstore 331 exceeds this number of bytes. Value is checked by a thread that runs 332 every hbase.server.thread.wakefrequency. 333 </description> 334 </property> 335 <property> 336 <name>hbase.hregion.preclose.flush.size</name> 337 <value>5242880</value> 338 <description> 339 If the memstores in a region are this size or larger when we go 340 to close, run a "pre-flush" to clear out memstores before we put up 341 the region closed flag and take the region offline. On close, 342 a flush is run under the close flag to empty memory. During 343 this time the region is offline and we are not taking on any writes. 344 If the memstore content is large, this flush could take a long time to 345 complete. The preflush is meant to clean out the bulk of the memstore 346 before putting up the close flag and taking the region offline so the 347 flush that runs under the close flag has little to do. 348 </description> 349 </property> 350 <property> 351 <name>hbase.hregion.memstore.block.multiplier</name> 352 <value>2</value> 353 <description> 354 Block updates if memstore has hbase.hregion.block.memstore 355 time hbase.hregion.flush.size bytes. Useful preventing 356 runaway memstore during spikes in update traffic. Without an 357 upper-bound, memstore fills such that when it flushes the 358 resultant flush files take a long time to compact or split, or 359 worse, we OOME. 360 </description> 361 </property> 362 <property> 363 <name>hbase.hregion.max.filesize</name> 364 <value>268435456</value> 365 <description> 366 Maximum HStoreFile size. If any one of a column families' HStoreFiles has 367 grown to exceed this value, the hosting HRegion is split in two. 368 Default: 256M. 369 </description> 370 </property> 371 <property> 372 <name>hbase.hstore.compactionThreshold</name> 373 <value>3</value> 374 <description> 375 If more than this number of HStoreFiles in any one HStore 376 (one HStoreFile is written per flush of memstore) then a compaction 377 is run to rewrite all HStoreFiles files as one. Larger numbers 378 put off compaction but when it runs, it takes longer to complete. 379 </description> 380 </property> 381 <property> 382 <name>hbase.hstore.blockingStoreFiles</name> 383 <value>7</value> 384 <description> 385 If more than this number of StoreFiles in any one Store 386 (one StoreFile is written per flush of MemStore) then updates are 387 blocked for this HRegion until a compaction is completed, or 388 until hbase.hstore.blockingWaitTime has been exceeded. 389 </description> 390 </property> 391 <property> 392 <name>hbase.hstore.blockingWaitTime</name> 393 <value>90000</value> 394 <description> 395 The time an HRegion will block updates for after hitting the StoreFile 396 limit defined by hbase.hstore.blockingStoreFiles. 397 After this time has elapsed, the HRegion will stop blocking updates even 398 if a compaction has not been completed. Default: 90 seconds. 399 </description> 400 </property> 401 <property> 402 <name>hbase.hstore.compaction.max</name> 403 <value>10</value> 404 <description>Max number of HStoreFiles to compact per 'minor' compaction. 405 </description> 406 </property> 407 <property> 408 <name>hbase.hregion.majorcompaction</name> 409 <value>86400000</value> 410 <description>The time (in miliseconds) between 'major' compactions of all 411 HStoreFiles in a region. Default: 1 day. 412 Set to 0 to disable automated major compactions. 413 </description> 414 </property> 415 <property> 416 <name>hbase.mapreduce.hfileoutputformat.blocksize</name> 417 <value>65536</value> 418 <description>The mapreduce HFileOutputFormat writes storefiles/hfiles. 419 This is the minimum hfile blocksize to emit. Usually in hbase, writing 420 hfiles, the blocksize is gotten from the table schema (HColumnDescriptor) 421 but in the mapreduce outputformat context, we don't have access to the 422 schema so get blocksize from Configuation. The smaller you make 423 the blocksize, the bigger your index and the less you fetch on a 424 random-access. Set the blocksize down if you have small cells and want 425 faster random-access of individual cells. 426 </description> 427 </property> 428 <property> 429 <name>hfile.block.cache.size</name> 430 <value>0.2</value> 431 <description> 432 Percentage of maximum heap (-Xmx setting) to allocate to block cache 433 used by HFile/StoreFile. Default of 0.2 means allocate 20%. 434 Set to 0 to disable. 435 </description> 436 </property> 437 <property> 438 <name>hbase.hash.type</name> 439 <value>murmur</value> 440 <description>The hashing algorithm for use in HashFunction. Two values are 441 supported now: murmur (MurmurHash) and jenkins (JenkinsHash). 442 Used by bloom filters. 443 </description> 444 </property> 445 <property> 446 <name>zookeeper.session.timeout</name> 447 <value>180000</value> 448 <description>ZooKeeper session timeout. 449 HBase passes this to the zk quorum as suggested maximum time for a 450 session. See http://hadoop.apache.org/zookeeper/docs/current/zookeeperProgrammers.html#ch_zkSessions 451 "The client sends a requested timeout, the server responds with the 452 timeout that it can give the client. " 453 In milliseconds. 454 </description> 455 </property> 456 <property> 457 <name>zookeeper.znode.parent</name> 458 <value>/hbase</value> 459 <description>Root ZNode for HBase in ZooKeeper. All of HBase's ZooKeeper 460 files that are configured with a relative path will go under this node. 461 By default, all of HBase's ZooKeeper file path are configured with a 462 relative path, so they will all go under this directory unless changed. 463 </description> 464 </property> 465 <property> 466 <name>zookeeper.znode.rootserver</name> 467 <value>root-region-server</value> 468 <description>Path to ZNode holding root region location. This is written by 469 the master and read by clients and region servers. If a relative path is 470 given, the parent folder will be ${zookeeper.znode.parent}. By default, 471 this means the root location is stored at /hbase/root-region-server. 472 </description> 473 </property> 474 <!-- 475 The following three properties are used together to create the list of 476 host:peer_port:leader_port quorum servers for ZooKeeper. 477 --> 478 <property> 479 <name>hbase.zookeeper.quorum</name> 480 <value>localhost</value> 481 <description>Comma separated list of servers in the ZooKeeper Quorum. 482 For example, "host1.mydomain.com,host2.mydomain.com,host3.mydomain.com". 483 By default this is set to localhost for local and pseudo-distributed modes 484 of operation. For a fully-distributed setup, this should be set to a full 485 list of ZooKeeper quorum servers. If HBASE_MANAGES_ZK is set in hbase-env.sh 486 this is the list of servers which we will start/stop ZooKeeper on. 487 </description> 488 </property> 489 <property> 490 <name>hbase.zookeeper.peerport</name> 491 <value>2888</value> 492 <description>Port used by ZooKeeper peers to talk to each other. 493 See http://hadoop.apache.org/zookeeper/docs/r3.1.1/zookeeperStarted.html#sc_RunningReplicatedZooKeeper 494 for more information. 495 </description> 496 </property> 497 <property> 498 <name>hbase.zookeeper.leaderport</name> 499 <value>3888</value> 500 <description>Port used by ZooKeeper for leader election. 501 See http://hadoop.apache.org/zookeeper/docs/r3.1.1/zookeeperStarted.html#sc_RunningReplicatedZooKeeper 502 for more information. 503 </description> 504 </property> 505 <!-- End of properties used to generate ZooKeeper host:port quorum list. --> 506 <!-- 507 Beginning of properties that are directly mapped from ZooKeeper's zoo.cfg. 508 All properties with an "hbase.zookeeper.property." prefix are converted for 509 ZooKeeper's configuration. Hence, if you want to add an option from zoo.cfg, 510 e.g. "initLimit=10" you would append the following to your configuration: 511 <property> 512 <name>hbase.zookeeper.property.initLimit</name> 513 <value>10</value> 514 </property> 515 --> 516 <property> 517 <name>hbase.zookeeper.property.initLimit</name> 518 <value>10</value> 519 <description>Property from ZooKeeper's config zoo.cfg. 520 The number of ticks that the initial synchronization phase can take. 521 </description> 522 </property> 523 <property> 524 <name>hbase.zookeeper.property.syncLimit</name> 525 <value>5</value> 526 <description>Property from ZooKeeper's config zoo.cfg. 527 The number of ticks that can pass between sending a request and getting an 528 acknowledgment. 529 </description> 530 </property> 531 <property> 532 <name>hbase.zookeeper.property.dataDir</name> 533 <value>${hbase.tmp.dir}/zookeeper</value> 534 <description>Property from ZooKeeper's config zoo.cfg. 535 The directory where the snapshot is stored. 536 </description> 537 </property> 538 <property> 539 <name>hbase.zookeeper.property.clientPort</name> 540 <value>2181</value> 541 <description>Property from ZooKeeper's config zoo.cfg. 542 The port at which the clients will connect. 543 </description> 544 </property> 545 <property> 546 <name>hbase.zookeeper.property.maxClientCnxns</name> 547 <value>30</value> 548 <description>Property from ZooKeeper's config zoo.cfg. 549 Limit on number of concurrent connections (at the socket level) that a 550 single client, identified by IP address, may make to a single member of 551 the ZooKeeper ensemble. Set high to avoid zk connection issues running 552 standalone and pseudo-distributed. 553 </description> 554 </property> 555 <!-- End of properties that are directly mapped from ZooKeeper's zoo.cfg --> 556 <property> 557 <name>hbase.rest.port</name> 558 <value>8080</value> 559 <description>The port for the HBase REST server.</description> 560 </property> 561 <property> 562 <name>hbase.rest.readonly</name> 563 <value>false</value> 564 <description> 565 Defines the mode the REST server will be started in. Possible values are: 566 false: All HTTP methods are permitted - GET/PUT/POST/DELETE. 567 true: Only the GET method is permitted. 568 </description> 569 </property> 570 </configuration>