干净虚拟机(centos 6.4)上从头到尾安装并调试Mdrill(二)

7、安装mdrill

服务安装

拷贝alimama-adhoc.tar.gz/home/mdrill,解压并替换lib目录中的hadoop-core-0.20.2.jarhadoop-core-0.20.2-cdh3u3.jar;拷贝hadoop服务lib目录下guava-r09-jarjar.jarmdril服务lib目录。

tar -xvf alimama-adhoc.tar.gz

cd /home/mdrill/alimama/adhoc-core/lib/

cp    $HADOOP_HOME/hadoop-core-0.20.2-cdh3u3.jar hadoop-core-0.20.2-cdh3u3.jar

rm hadoop-core-0.20.2.jar

cp $HADOOP_HOME/lib/guava-r09-jarjar.jar   /home/mdrill/alimama/adhoc-core/lib/

服务配置

配置storm.yaml文件

vi /home/mdrill/alimama/adhoc-core/conf/storm.yaml

修改完成后内容如下,主要是对主机地址,shards个数,worker启动参数进行了调整:


####zookeeper配置####

 storm.zookeeper.servers:

       - "mdrill"

 storm.zookeeper.port: 2181

 storm.zookeeper.root: "/higo2"

####蓝鲸配置####

storm.local.dir: "/home/mdrill/alimama/bluewhale/stormwork"

nimbus.host: "mdrill"

####hadoop配置####

 hadoop.conf.dir:   "/home/mdrill/hadoop-0.20.2-cdh3u3/conf"

 hadoop.java.opts: "-Xmx128m"

 ####mdrill存储目录配置####

 higo.workdir.list:   "/home/mdrill/alimama/higoworkerdir"

 #----mdrill的表格列表在hdfs下的路径-----

 higo.table.path:   "/mdrill/tablelist"

 #----mdrill中启动的solr使用的初始端口号-----

 higo.solr.ports.begin: 51110

 #----mdrill分区方式,目前支持default,day,month,singledefault是将一个月分成3个区,single意味着没有分区-----

 higo.partion.type: "month"

 #----创建索引生成的每个shard的并行----

 higo.index.parallel: 2

 #----启动的shard的数,每个shard为一个solr实例,结合cpu个数和内存进行配置,1048G内存配置60----

 higo.shards.count: 2

 #----基于冗余的ha,设置为1表示没有冗余,如果设置为2,则冗余号位0,1----

 higo.shards.replication: 1

 #----启动的merger serverworker数量,建议根据机器数量设定----

 higo.mergeServer.count: 1

 #----mdrill同时最多加载的分区个数,取决于内存与数据量----

 higo.cache.partions: 1

 #----通用的workerjvm参数配置----

 worker.childopts: "-Xms128m -Xmx128m   -Xmn64m -XX:SurvivorRatio=2 -XX:PermSize=48m -XX:MaxPermSize=64m   -XX:+UseParallelGC -XX:ParallelGCThreads=2 -XX:+UseAdaptiveSizePolicy   -XX:+PrintGCDetails -XX:+PrintGCTimeStamps    -Xloggc:%storm.home%/logs/gc-%port%.log "

 #----个别的worker的端口的jvm参数配置,通常用于给merger server单独配置----

 worker.childopts.6601: "-Xms128m   -Xmx128m -Xmn64m -XX:SurvivorRatio=2 -XX:PermSize=48m -XX:MaxPermSize=64m   -XX:+UseParallelGC -XX:ParallelGCThreads=2 -XX:+UseAdaptiveSizePolicy   -XX:+PrintGCDetails -XX:+PrintGCTimeStamps    -Xloggc:%storm.home%/logs/gc-%port%.log "

 worker.childopts.6602: "-Xms128m   -Xmx128m -Xmn64m -XX:SurvivorRatio=2 -XX:PermSize=48m -XX:MaxPermSize=64m   -XX:+UseParallelGC -XX:ParallelGCThreads=2 -XX:+UseAdaptiveSizePolicy   -XX:+PrintGCDetails -XX:+PrintGCTimeStamps    -Xloggc:%storm.home%/logs/gc-%port%.log "

 #----任务分配,注意,原先分布在同一个位置的任务,迁移后还必须分布在一起,比如说acker:0merger:10分布在一起了,那么迁移后他们还要在一起,为了保险起见,建议同一个端口跑多个任务的,可以像下面ackerheartbeat那样----

 mdrill.task.ports.adhoc:   "6901~6902"

 mdrill.task.assignment.adhoc:

       #- "####看到这里估计大家都会晕,但是这个是任务调度很重要的地方,出去透透气,回来搞定这里吧####"    

       #- "####注解:merge表示是merger server;shard@0表示是shard的第0个冗余,__ackerheartbeat是内部线程,占用资源很小####"    

       #- "####下面为初始分布,用于同一台机器之间没有宕机的调度####"

       #- "####切记,一个端口只能分配一个shard,千万不要将mergeshardshardshard分配到同一个端口里,这会引起计算数据的不准确的####"

       #- "merge:10&shard@0:0~4;tiansuan1.kgbNaN4:6601,6701~6705;6602,6711~6715;6603,6721~6725"

       #-   "merge:1&shard@0:5~9;tiansuan2.kgbNaN4:6601,6701~6705;6602,6711~6715;6603,6721~6725"

       #-   "merge:2&shard@0:48~53;adhoc1.kgbNaN6:6601,6701~6706;6602,6711~6716;6603,6721~6726"

       #- "merge:3&shard@0:10~14;adhoc2.kgbNaN6:6601,6701~6705;6602,6711~6715;6603,6721~6725"

       #-   "merge:4&shard@0:15~19;adhoc3.kgbNaN6:6601,6701~6705;6602,6711~6715;6603,6721~6725"

       #-   "merge:5&shard@0:20~24;adhoc4.kgbNaN6:6601,6701~6705;6602,6711~6715;6603,6721~6725"

     #-   "merge:6&shard@0:25~30;adhoc5.kgbNaN6:6601,6701~6706;6602,6711~6716;6603,6721~6726"

       #-   "merge:7&shard@0:31~36;adhoc6.kgbNaN6:6601,6701~6706;6602,6711~6716;6603,6721~6726"

       #-   "merge:8&shard@0:37~41;adhoc7.kgbNaN6:6601,6701~6705;6602,6711~6715;6603,6721~6725"

       #-   "merge:9&shard@0:42~47;adhoc8.kgbNaN6:6601,6701~6706;6602,6711~6716;6603,6721~6726"

       - "shard@0:0~1;mdrill:6701~6702;6711~6712;6721~6722"

       - ""

       - "##__acker,heartbeat,merge:0他们共用同一个端口,记住要时时刻刻,他们共享的都是同一个端口,别拆开了##"

       - "__acker:0;mdrill:6601;6602;6603"

       - "heartbeat:0;mdrill:6601;6602;6603"

       - "merge:0;mdrill:6601;6602;6603"

       - ""

       -   "#######################################################################################################################"

       #- "####如果某台机器宕机了,任务无法在固定的分配都某一台,就需要使用下面的配置,宕机的任务会迁移到新的机器上####"

       #- "####注意这里的端口不要与上面的端口有重叠,某个任务也不要继续分配在上面的机器,因为那台机器已经宕机了####"

       #- "####要注意如果某一台机器宕机,他的任务要均匀的分布到集群其他机器上,不要全部迁移到同一台机器上####"

       #- "merge:2&shard@0:58,59,16,26,38;tiansuan1.kgbNaN4:6604,6721~6725;6605,6731~6735"

       #-   "merge:3&shard@0:59,50,17,27,39;tiansuan2.kgbNaN4:6604,6721~6725;6605,6731~6735"

       #-   "merge:4&shard@0:1,10,18,28,40,57;adhoc1.kgbNaN6:6604,6721~6726;6605,6731~6736"

       #- "shard@0:2,51,19,29,41;adhoc2.kgbNaN6:6604,6721~6725;6605,6731~6735"

       #-   "merge:6&shard@0:3,52,20,30,42;adhoc3.kgbNaN6:6604,6721~6725;6605,6731~6735"

       #-   "merge:7&shard@0:4,53,33,31,43;adhoc4.kgbNaN6:6604,6721~6725;6605,6731~6735"

       #- "merge:8&shard@0:6,11,21,32,44,56;adhoc5.kgbNaN6:6604,6721~6726;6605,6731~6736"

       #-   "merge:9&shard@0:7,12,22,37,45,55;adhoc6.kgbNaN6:6604,6721~6726;6605,6731~6736"

       #-   "merge:5&shard@0:8,13,23,34,46;adhoc7.kgbNaN6:6604,6721~6725;6605,6731~6735"

       #- "merge:10&shard@0:9,14,24,35,54,5,;adhoc8.kgbNaN6:6604,6721~6726;6605,6731~6736"

       #-   "merge:1&shard@0:15,25,36,47,0;adhocbak.kgbNaN6:6604,6721~6726;6605,6731~6736"

       #- "##__ackerheartbeat,merge:0他们共用同一个端口,记住要时时刻刻,他们共享的都是同一个端口##"

       #- "__acker:0;adhoc2.kgbNaN6:6604;6605"

       #- "heartbeat:0;adhoc2.kgbNaN6:6604;6605"

       #- "merge:0;adhoc2.kgbNaN6:6604;6605"

 #----mdrill个别表的分区设定,这里为rpt_seller_all_m----

 higo.mode.fact_wirelesspv_clickcosteffect1_lbs_adhoc_d:   "@nothedate@"

 higo.partion.type.fact_p4ppvclick_qrycatpid_d:   "day"

 higo.partion.type.rpt_seller_all_m:   "month"

 higo.partion.type.r_rpt_cps_luna_item:   "default"

 higo.partion.type.fact_wirelesspv_clickcosteffect1_lbs_adhoc_d:   "day"

 higo.partion.type.r_rpt_cps_adhoc_seller:   "default"

 higo.partion.type.rpt_seller_all_m_single:   "single"

 higo.partion.type.rpt_seller_all_m_dir:   "dir@thedate"

 higo.partion.type.ods_allpv_ad_d:   "day"

 higo.partion.type.fact_wirelesspv_clickcosteffect1_app_adhoc_d_1:   "day"

 higo.mode.fact_wirelesspv_clickcosteffect1_app_adhoc_d_1:   "@nonzipout@nothedate@"

 higo.partion.type.fact_wirelesspv_clickcosteffect1_app_adhoc_d_2:   "day"

 higo.mode.fact_wirelesspv_clickcosteffect1_app_adhoc_d_2:   "@nonzipout@nothedate@"

 higo.partion.type.fact_wirelesspv_clickcosteffect1_app_adhoc_d_3:   "day"

 higo.mode.fact_wirelesspv_clickcosteffect1_app_adhoc_d_3:   "@nonzipout@nothedate@"

 higo.partion.type.fact_wirelesspv_clickcosteffect1_app_adhoc_d_4:   "day"

 higo.mode.fact_wirelesspv_clickcosteffect1_app_adhoc_d_4:   "@nonzipout@nothedate@"

 #---注意只有分区类型为day的时候才能配置nothedate参数----

 higo.partion.type.fact_wirelesspv_clickcosteffect1_adhoc_d:   "day"

 higo.mode.fact_wirelesspv_clickcosteffect1_adhoc_d:   "@nonzipout@nothedate@"

 higo.partion.type.fact_wireless_pvcost_clickeffect_zz_adhoc_d:   "day"

 higo.mode.fact_wireless_pvcost_clickeffect_zz_adhoc_d:   "@nonzipout@nothedate@" 

 #-------------hdfs mode-----------------

 higo.partion.type.s_ods_log_p4pclick:   "day"

 higo.mode.s_ods_log_p4pclick:   "@fdt@nonzipout@nothedate@"

 higo.mode.ods_allpv_ad_d:   "@hdfs@fdt@sigment:15@blocksize:1073741824@iosortmb:256@igDataChange@"  

 higo.index.parallel.ods_allpv_ad_d: 40

 #-------------为个别表单独指定并行-----------------

 higo.index.parallel.r_rpt_cps_luna_item: 10

 ####--实时模式相关配置--####

 #----binlog记录的位置,默认为local如果hdfs稳定可以写hdfs----

 higo.realtime.binlog: "local"

 #----tt实时导入到mdrill----

 higo.partion.type.rpt_quanjing_tanxclick:   "default"

 higo.partion.type.rpt_quanjing_tanxpv:   "default"

 higo.partion.type.s_ods_mm_adzones:   "default"

 higo.mode.s_ods_mm_adzones:   "@fieldcontains:status=A@"

 higo.mode.tanx_pv:   "@realtime@cleanPartions:99@"

 higo.partion.type.tanx_pv: "day"

 tanx_pv-log: "tanx_pv"

 tanx_pv-subid: "01061423317TVV8KSN"

 tanx_pv-accesskey:   "604a3330-5c1f-4488-9466-de63f08fe297"

 tanx_pv-parse:   "com.alimama.quanjingmonitor.mdrillImport.parse.tanx_pv"

 tanx_pv-reader:   "com.alimama.quanjingmonitor.mdrillImport.reader.TT4Reader"

 tanx_pv-commitbatch: 200

 tanx_pv-commitbuffer: 2000

 tanx_pv-spoutbuffer: 3000

 tanx_pv-boltbuffer: 500

 tanx_pv-boltIntervel: 1000

 tanx_pv-spoutIntervel: 1000

 tanx_pv-threads: 32

 tanx_pv-threads_reduce: 12

 tanx_pv-mode: "merger"

 tanx_pv-timeoutSpout: 60

 tanx_pv-timeoutBolt: 30

 tanx_pv-timeoutCommit: 100

 higo.mode.tanx_click:   "@realtime@cleanPartions:99@"

 higo.partion.type.tanx_click:   "day"

 tanx_click_zhitou-log:   "tanx_click_zhitou"

 tanx_click_zhitou-subid:   "0106143856X2HG2CRT"

 tanx_click_zhitou-accesskey:   "9f126ea7-e728-464f-92db-8fe03aed561a"

 tanx_click_zhitou-parse:   "com.alimama.quanjingmonitor.mdrillImport.parse.tanx_click_zhitou"

 tanx_click_zhitou-reader:   "com.alimama.quanjingmonitor.mdrillImport.reader.TT4Reader"

 tanx_click_zhitou-commitbatch: 200

 tanx_click_zhitou-commitbuffer: 2000

 tanx_click_zhitou-spoutbuffer: 500

 tanx_click_zhitou-boltbuffer: 500

 tanx_click_zhitou-boltIntervel: 10

 tanx_click_zhitou-spoutIntervel: 100

 tanx_click_zhitou-threads: 2

 tanx_click_zhitou-threads_reduce: 12

 tanx_click_zhitou-mode: "merger"

 tanx_click_zhitou-timeoutSpout: 30

 tanx_click_zhitou-timeoutBolt: 20

 tanx_click_zhitou-timeoutCommit: 60

 tanx_click_rt-log: "tanx_rd_log"

 tanx_click_rt-subid:   "0106144243FH80E8QP"

 tanx_click_rt-accesskey:   "a53f9bb3-26c0-4ba8-9389-13057b9e5d34"

 tanx_click_rt-parse:   "com.alimama.quanjingmonitor.mdrillImport.parse.tanx_click_rt"

 tanx_click_rt-reader:   "com.alimama.quanjingmonitor.mdrillImport.reader.TT4Reader"

 tanx_click_rt-commitbatch: 200

 tanx_click_rt-commitbuffer: 2000

 tanx_click_rt-spoutbuffer: 500

 tanx_click_rt-boltbuffer: 500

 tanx_click_rt-boltIntervel: 10

 tanx_click_rt-spoutIntervel: 100

 tanx_click_rt-threads: 2

 tanx_click_rt-threads_reduce: 12

 tanx_click_rt-mode: "merger"

 tanx_click_rt-timeoutSpout: 30

 tanx_click_rt-timeoutBolt: 20

 tanx_click_rt-timeoutCommit: 60

 tanx_user_rt-log: "user_rd"

 tanx_user_rt-subid:   "0106144909UBGMF3IB"

 tanx_user_rt-accesskey:   "91bb5ad9-dc37-49b1-be75-8e1de52e4407"

 tanx_user_rt-parse:   "com.alimama.quanjingmonitor.mdrillImport.parse.tanx_user_rt"

 tanx_user_rt-reader:   "com.alimama.quanjingmonitor.mdrillImport.reader.TT4Reader"

 tanx_user_rt-commitbatch: 200

 tanx_user_rt-commitbuffer: 2000

 tanx_user_rt-spoutbuffer: 500

 tanx_user_rt-boltbuffer: 500

 tanx_user_rt-boltIntervel: 10

 tanx_user_rt-spoutIntervel: 100

 tanx_user_rt-mode: "merger"

 tanx_user_rt-threads: 2

 tanx_user_rt-threads_reduce: 12

 tanx_user_rt-timeoutSpout: 30

 tanx_user_rt-timeoutBolt: 20

 tanx_user_rt-timeoutCommit: 60

 mdrill.task.ports.p4p_pv2topology:   "6801~6806"

 mdrill.task.assignment.p4p_pv2topology:

       - "random"

 

 ####蓝鲸各种心跳间隔配置-一般不需要改动####

 nimbus.task.timeout.secs: 30

 task.heartbeat.frequency.secs: 6

 nimbus.supervisor.timeout.secs: 300

 supervisor.heartbeat.frequency.secs: 20

 nimbus.monitor.freq.secs: 60

 supervisor.worker.timeout.secs: 600

 supervisor.monitor.frequency.secs: 90

 worker.heartbeat.frequency.secs: 30

 #----每台机器启动的worker进程端口列表,一般请不要改动----

 supervisor.slots.ports:

      - 6601

      - 6602

      - 6603

      - 6604

      - 6605

      - 6606

      - 6701

      - 6702

      - 6703

      - 6704

      - 6705

      - 6706

      - 6707

      - 6708

      - 6709

      - 6710

      - 6711

      - 6712

      - 6713

      - 6714

      - 6715

      - 6716

      - 6717

      - 6718

      - 6719

      - 6720

      - 6721

      - 6722

      - 6723

      - 6724

      - 6725

      - 6726

      - 6727

      - 6728

      - 6729

      - 6730

      - 6731

      - 6732

      - 6733

      - 6734

      - 6735

      - 6736

      - 6737

      - 6738

      - 6739

      - 6801

      - 6802

      - 6803

      - 6804

      - 6805

      - 6806

      - 6901

      - 6902

####以下为adhoc项目专用,跟mdrill无关,不启动UI的话可忽略####

 higo.ui.conf.dir:   "/home/taobao/bluewhale/ui16/alimama/adhoc-core/webapp"

 higo.download.offline.conn:   "jdbc:mysql://tiansuan1.kgbNaN4:3306/adhoc_download?useUnicode=true&characterEncoding=utf8"

 higo.download.offline.username:   "adhoc"

 higo.download.offline.passwd:   "adhoc"

 higo.download.offline.store: "/group/tbdp-etao-adhoc/p4padhoc/download/offline"

 mysql.url:   "jdbc:mysql://tiansuan1.kgbNaN4:3306/query_analyser?useUnicode=true&characterEncoding=UTF-8"

 mysql.username: "adhoc"

 mysql.password: "adhoc"

 dev.name.list:

     - "木晗"

     - "子落"

     - "张壮"

     - "行列"

由于是在配置单节点学习环境,以上配置去除了shards的备份策略。

新建higoworkerdir

mkdir /home/mdrill/alimama/higoworkerdir

 

8、启动和退出

8.1启动hadoop

start-all.sh

8.2启动zookeeper

zkServer.sh start

8.3启动nimbus

cd /home/mdrill/alimama/adhoc-core/bin/

chmod 777 ./bluewhale

nohup ./bluewhale nimbus >nimbus.log   &

8.4启动supervisor

cd /home/mdrill/alimama/adhoc-core/bin/

nohup ./bluewhale supervisor   >supervisor.log &

注意:每次启动前需要清理 /home/mdrill/alimama/bluewhale/stormwork文件夹。

8.5启动uijdbc接口

cd  /home/mdrill/alimama/adhoc-core/bin/

mkdir ./ui

nohup ./bluewhale   mdrillui 1107 ../lib/adhoc-web-0.18-beta.jar ./ui >ui.log &

启动后,可以通过浏览器打开mdrill1107端口,看是否能正常打开即可,可看到下图:

干净虚拟机(centos 6.4)上从头到尾安装并调试Mdrill(二)_第1张图片

所有服务启动完成后,可以看到如下进程:

[mdrill@mdrill bin]$ jps

62621 Jps

60200 NameNode

60467 JobTracker

61568 Supervisor

60401 SecondaryNameNode

61887 MdrillUi

61329 NimbusServer

60592 TaskTracker

60300 DataNode

35049 QuorumPeerMain

未完待续

你可能感兴趣的:(mdrill)