1、hbase.roo1.dirhdfsc1.uster1.dghbasehbasc.c1.uster,distributcdtruehbase.zookeeper,quorumhadoopNode1,hadoopNode2,hadoopNode3,hadoopNode4.hadoopNode5hbase.zookeeper,property.dataDirhomeambowzkdatahdata6 .配Jtregionserver（配置每一个机器名子节点名不要配主节名）在hbaseCOnf/下新建regionserver文件，添加如入内容hadoopNodc3hadoopNode4hadoopN
2、ode57 .scp-rhbase到其他节点amboWWhadOOPNode1.confSscp-rapphbase-1. 3.2ambowhadoopNode5：appambowhadoopNodc1confSscp-rapphbasc-1.3.2ambowhadoopNode4：vappambowhadoopNode1.confSscprapphbase-1.3.2ambowhadoopNodc3：appambHr0hadoopNode1confSscp-rapphbase-1.3.2ambowhadoopNode2：Vapp/ambowhadoopNode1.confSscp/.bash
3、profi1.eambow0hadoopNodc5:amboWWhadoOPNode1.confSscp/.bash_profi1.eambowhadoopNode4：ambowhadoopNodc1confSscp/,bash_profi1eambowhadoopNode3：ambowhadoopNode1confJSscp/.bashprofi1.eambowQhadoopNodc2:各节点重新加载：sourceV.bash_profiIe启动hdfsstart-dfs.sh六、F1.ume安装6.1 安装1 .解压tar-ZXVfaPaChc-fIumcT.6.O-bin.tar.gza
4、mbowhadoopNode3f1.umeT.6.0$tar-zxvfapache-fIume-1. 6.O-bin.tar.gz-C7app2 .然后进入f1.ume的目录，修改Conf下的f1.umc-cnv.sh,配置JAVAJIOMEexportJVAJIOME=homeambowappjdk1.8.O1213 .配置.bash-ProfiIe文件exportF1.UME_HOME=/home/ambowapp/fIume-1.6.OexprotPATH=$PATH:$F1.UMEJioME/bin七、Kafka安装：7.1安装I.下载Apachekafka官方：http：/kafka
5、apache,o1.wn1.oads.htm1.Sca1.a2.11-kafka_2.11-0.10.2.0.tgz(asc,md5)注:当SCaIa用的是2.11那Kafka选择kafka_2.11-0.102.O.10.2才是Kafka的版本Kafka集群安装：1 .安装JDK&配置JAVAIOME2 .安装Zookeeper参照ZOokeCPCr官网搭建一个ZK集群，并启动ZK臬群。3 .解压Kafka安装包ambowhadoopNode1ambowStar-zxvfkafka_2.11一0.10.2.1.tgz-CVapp/4.配置环境变量exportKKAH()ME-homcamb
6、owapp/kafka2.11-0.10.2.1exportPTH=PTH:$KAFKAK)ME/bin5.修改配置文件config/server.propertiesviserver,properties#为依次增长的：0、1、2、3、4,集群中节点唯一idbroker,id-0#删除主题的配置，默认是fa1.se生产环境设为fa1.sede1.ete,topic,enab1.e=true#监听的主机及端口号各节点改为本机相应的hostName1isteners=P1.AINTEXT：/hadoopNode1：9092邓afka的消息数据存储路径1.og.dirs-7homcambowkaf
7、ka1.)ataIogs#创建主题的时候，默认有1个分区num.partitions=3#指定ZooKeeper集群列表，各节点以逗号分zookeeper,connect-hadoopode1.:2181,hadoopNode2:2181,hadoopNode3:2181,hadoopNode4：2181.hadoopNode5:21816 .分发至各个节点ambowhadoopNode1app$scp-rkafka2.11-0.10.2.1ambowhadoopNode5：VappambowhadoopNode1app$scp-rkafka2.11-0.10.2.1ambowhadoopNo
8、de4：appambowhadoopNode1app$scp-rkafka_2.11-0.10.2.1ambowhadoopNode3：VappambowhadoopNode1app$scp-rkafka2.11-0.10.2.1ambowhadoopNode2：VappambowhadoopNode1app$scp-r7.bash_profi1.eambowMIadoOPNode5：、ambowhadoopNode1app1$scp-r、/.bashprofiIcambowhadoopNodc4:ambowhadoopNode1app：$scp-r.bash_profiIeambowWhad
9、OOPNode3：,、ambowhadoopNode1app：$scp-r.bash_profi1.eambowhadoopNode2：source/.bash_profiIe7 .修改各个节点的配置文件：为依次增长的：0、1、2、3、4,集群中节点唯一idbroker,id-0力监听的主机及端口号各节点改为本机相应的hos1.Name1isteners-P1.AINTEXThadoopNode1：90928 .各台节点上启动Kafka服务ambowhadoopNode1app$kafka-server-start.sh$KAFK.*1.1.OME/config/server.properties&注：要先启动各节点ZOOkeePerZkServ1.et.shstartambowhadoopNodc1app$kafka-SCrVer-Stop.sh
