hbase centOS生产环境配置笔记 (1 NameNode, 1 ResourceManager, 3 DataNode)

本次是第一次在生产环境部署HBase,本文若有配置上的不妥之处还请高手指正。

hadoop版本:hadoop-2.4.1

HBase版本:hbase-0.98.6.1-hadoop2

Zookeeper版本:ZooKeeper3.4.6

JDK:1.6

操作系统:centOS6.3 64bit

 

1. 安装snappy

#yum install snappy
Installed: snappy.x86_64 0:1.1.0-1.el6
#yum install snappy-devel

记录一下,当前GCC版本4.4.7

 

2. 由于hadoop官网下载的bin包中包含的native lib是32bit环境编译的,需要重新编译hadoop源码以获得64bit的native lib。

$ file libhadoop.so.1.0.0
libhadoop.so.1.0.0: ELF 32-bit LSB shared object, Intel 80386, version 1 (SYSV), dynamically linked, not stripped

(1) 在hadoop官网下载 hadoop-2.4.1-src.tar.gz。解压后打开 BUILDING.txt

Requirements:

* Unix System
* JDK 1.6+
* Maven 3.0 or later
* Findbugs 1.3.9 (if running findbugs)
* ProtocolBuffer 2.5.0
* CMake 2.6 or newer (if compiling native code)
* Internet connection for first build (to fetch all Maven and Hadoop dependencies)

按照上述指示,

(2) 下载 Maven:http://maven.apache.org/download.cgi

(3) 下载 ProtocolBuffer 2.5.0: https://code.google.com/p/protobuf/downloads/list,解压编译

$ cd protobuf-2.6.1
$ ./configure

出现以下错误:

configure: error: C++ preprocessor "/lib/cpp" fails sanity check

经过网上搜索,需要安装c++相关的库:

# yum install glibc-headers
# yum install gcc-c++

再次执行 ./configure,这次通过了。

然后就可以make了

$ make
# make install

安装成功后,可以在/usr/local/lib/下看到库文件:

-rw-r--r-- 1 root root 22955150 Oct 23 11:41 libprotobuf.a
-rwxr-xr-x 1 root root      988 Oct 23 11:41 libprotobuf.la
-rw-r--r-- 1 root root  2233428 Oct 23 11:41 libprotobuf-lite.a
-rwxr-xr-x 1 root root     1023 Oct 23 11:41 libprotobuf-lite.la
lrwxrwxrwx 1 root root       25 Oct 23 11:41 libprotobuf-lite.so -> libprotobuf-lite.so.9.0.1
lrwxrwxrwx 1 root root       25 Oct 23 11:41 libprotobuf-lite.so.9 -> libprotobuf-lite.so.9.0.1
-rwxr-xr-x 1 root root  1038614 Oct 23 11:41 libprotobuf-lite.so.9.0.1
lrwxrwxrwx 1 root root       20 Oct 23 11:41 libprotobuf.so -> libprotobuf.so.9.0.1
lrwxrwxrwx 1 root root       20 Oct 23 11:41 libprotobuf.so.9 -> libprotobuf.so.9.0.1
-rwxr-xr-x 1 root root  9300568 Oct 23 11:41 libprotobuf.so.9.0.1
-rw-r--r-- 1 root root 39008232 Oct 23 11:41 libprotoc.a
-rwxr-xr-x 1 root root     1004 Oct 23 11:41 libprotoc.la
lrwxrwxrwx 1 root root       18 Oct 23 11:41 libprotoc.so -> libprotoc.so.9.0.1
lrwxrwxrwx 1 root root       18 Oct 23 11:41 libprotoc.so.9 -> libprotoc.so.9.0.1
-rwxr-xr-x 1 root root 13252621 Oct 23 11:41 libprotoc.so.9.0.1

 

(4) 安装cmake

# yum install cmake

 

(5) 安装openssl 

# yum install openssl
# yum install openssl-devel

 

(6) 安装zlib相关

# yum install zlib
# yum install zlib-devel

 

(7) 下载 ant 1.9.4

下载 http://ant.apache.org/bindownload.cgi

也可以直接 yum install ant

 

(8) 设置环境变量: 

export JAVA_HOME=/usr/java/jdk1.6.0_43
export PATH="$JAVA_HOME/bin:$PATH"

export MAVEN_HOME=/data2/hadoop_source/apache-maven-3.2.3
export PATH="$MAVEN_HOME/bin:$PATH"
export ANT_HOME=/data2/hadoop_source/apache-ant-1.9.4
export PATH="$PATH:$ANT_HOME/bin"

 

(9) 编译hadoop的native库:

/data2/hadoop_source/apache-maven-3.2.3/bin/mvn package -X -Pdist,native -Dtar -DskipTests -Drequire.snappy

 最后即使有部分project编译出错也没有关系,只要 hadoop-hdfs-project,hadoop-common-project 编译通过就可以了。

 (10) 编译完成后,可以找到native lib:

# find . -name 'lib*.so*'
./hadoop-hdfs-project/hadoop-hdfs/target/hadoop-hdfs-2.4.1/lib/native/libhdfs.so
./hadoop-hdfs-project/hadoop-hdfs/target/hadoop-hdfs-2.4.1/lib/native/libhdfs.so.0.0.0
./hadoop-hdfs-project/hadoop-hdfs/target/native/target/usr/local/lib/libhdfs.so
./hadoop-hdfs-project/hadoop-hdfs/target/native/target/usr/local/lib/libhdfs.so.0.0.0
./hadoop-common-project/hadoop-common/target/hadoop-common-2.4.1/lib/native/libhadoop.so.1.0.0
./hadoop-common-project/hadoop-common/target/hadoop-common-2.4.1/lib/native/libhadoop.so
./hadoop-common-project/hadoop-common/target/native/target/usr/local/lib/libhadoop.so.1.0.0
./hadoop-common-project/hadoop-common/target/native/target/usr/local/lib/libhadoop.so

将库文件拷贝至系统目录 /usr/local/lib64/

cp ./hadoop-hdfs-project/hadoop-hdfs/target/hadoop-hdfs-2.4.1/lib/native/* /usr/local/lib64/
cp ./hadoop-common-project/hadoop-common/target/hadoop-common-2.4.1/lib/native/* /usr/local/lib64/

进入/usr/local/lib64/,创建symbol link

# ln -s libhdfs.so.0.0.0 libhdfs.so
# ln -s libhadoop.so.1.0.0 libhadoop.so

 

(11) 下载hbase, http://mirrors.cnnic.cn/apache/hbase/hbase-0.98.7/

解压后进入目录,确认native lib是否已经安装成功:

$ LD_LIBRARY_PATH=/usr/local/lib64 ./bin/hbase org.apache.hadoop.util.NativeLibraryChecker

log4j:WARN No appenders could be found for logger (org.apache.hadoop.util.NativeCodeLoader).
log4j:WARN Please initialize the log4j system properly.
log4j:WARN See http://logging.apache.org/log4j/1.2/faq.html#noconfig for more info.
Native library checking:
hadoop: true /usr/local/lib64/libhadoop.so.1.0.0
zlib: true /lib64/libz.so.1
snappy: true /usr/lib64/libsnappy.so.1
lz4: true revision:99
bzip2: false

由于打算仅使用snappy作为压缩算法,所以仅仅需要确认 hadoop, snappy 是true就行了。

本次编译native lib过程中,出现了各种错误,都是因为上面列出的某些软件未安装。

 

无须在每个机器上再次编译,只需将编译过程中产生的那些 .a, .so, 文件拷贝至每个机器即可。当然了,symbol link还是要重新创建的。

---------------------------- 分割线 上方是 hadoop native lib 的编译 ------------------------------------------

 3. 配置hbase分布式环境的总体流程

本次安装部署的服务器为5台,hbase-0为NameNode, hbase-r为ResourceManager,hbase-1、hbase-2、hbase-3为DataNode。

所有服务器的安装的目录为:

hadoop程序目录:/hbase/hadoop

hbase程序目录:/hbase/hbase

zookeeper的程序目录:/hbase/zookeeper

hdfs数据目录(data node用):/hbase/hdfs

hdfs元数据目录(name node用):/hbase/hdfsmeta

hdfs临时目录:/hbase/hdfstmp

yarn中间数据目录:/hbase/yarnlocal

yarn日志目录:/hbase/yarnlog

job history server的临时目录:/hbase/mr-history/tmp

job history server的done目录:/hbase/mr-history/done

各个deameon的pid目录(安全起见,目录权限700):/hbase/var/pid

zookeeper的数据目录:/hbase/zookeeperdata

zookeeper的log目录:/hbase/zookeeperlog

hbase临时目录:/hbase/hbasetmp

 

ZooKeeper的下载地址:http://mirrors.cnnic.cn/apache/zookeeper/stable/

本次下载的是 ZooKeeper3.4.6。

本次安装采用独立的ZooKeeper,不使用hbase自带的ZooKeeper。

 

 

(1) 配置host name, /etc/hosts,为了方便使用,我将hostname配置成了以下样子:

10.161.150.10 hbase-0
10.161.150.20 hbase-r
10.161.150.11 hbase-1
10.161.150.12 hbase-2
10.161.150.13 hbase-3 

为了方便,还添加了1个 hbase-me,代表当前服务器的内网IP。

注意:上述配置hostname要放在靠前的位置,hbase的Master启动的时候,会根据ip查找hostname,相同ip的有多个hostname的时候,排在第1个的会被采用,然后被传播到RegionServer。

同时,将主机hostname修改成对应的名字。修改/etc/sysconfig/network,或者立即生效:

hostname hbase-0

 

(2) 配置免密码ssh

参考我的另一篇文章:http://www.cnblogs.com/got-my-way/p/4030923.html

(3) 解压hadoop包。

(4) 解压hbase包。

(5) 为了方便运行,将需要的环境变量配置在当前用户的.bashrc

vi ~/.bashrc

如下:

# For hadoop and hbase -------------------------------------
export JAVA_HOME="/usr/java/jdk1.6.0_43"
export PATH="$JAVA_HOME/bin:$PATH"
export LD_LIBRARY_PATH="$LD_LIBRARY_PATH:/usr/local/lib64:/usr/lib64:/usr/local/lib"
export HADOOP_PREFIX="/hbase/hadoop"
export HADOOP_HOME="/hbase/hadoop"
export HADOOP_YARN_HOME=$HADOOP_PREFIX
export HADOOP_CONF_DIR="/hbase/hadoop/etc/hadoop"
export HBASE_HOME="/hbase/hbase"
export HBASE_CONF_DIR="/hbase/hbase/conf"
export ZOOKEEPER_HOME="/hbase/zookeeper"
#log dir -----
export HADOOP_LOG_DIR="/hbase/var/log"
export YARN_LOG_DIR="/hbase/var/log"
#pid dir -----
export HADOOP_PID_DIR="/hbase/var/pid"
export HADOOP_SECURE_DN_PID_DIR="/hbase/var/pid"
export HADOOP_MAPRED_PID_DIR="/hbase/var/pid"
export YARN_PID_DIR="/hbase/var/pid"
#java options -----
export HADOOP_JVM_GC_OPTS="-XX:PermSize=512m -XX:MaxPermSize=512m -XX:+DisableExplicitGC -XX:SurvivorRatio=1 -XX:+UseConcMarkSweepGC -XX:+UseParNewGC -XX:+CMSParallelRemarkEnabled -XX:+UseCMSCompactAtFullCollection -XX:CMSFullGCsBeforeCompaction=0 -XX:+CMSClassUnloadingEnabled -XX:LargePageSizeInBytes=128M -XX:+UseFastAccessorMethods -XX:+UseCMSInitiatingOccupancyOnly -XX:CMSInitiatingOccupancyFraction=80 -XX:SoftRefLRUPolicyMSPerMB=0"
export HADOOP_JVM_SECURITY_OPTS="-Djava.security.krb5.realm=OX.AC.UK -Djava.security.krb5.kdc=kdc0.ox.ac.uk:kdc1.ox.ac.uk"
export HADOOP_NAMENODE_OPTS="${HADOOP_JVM_SECURITY_OPTS} ${HADOOP_JVM_GC_OPT}"
export HADOOP_DATANODE_OPTS="${HADOOP_JVM_SECURITY_OPTS} ${HADOOP_JVM_GC_OPT}"
export HADOOP_SECONDARYNAMENODE_OPTS="${HADOOP_JVM_SECURITY_OPTS} ${HADOOP_JVM_GC_OPT}"
export YARN_RESOURCEMANAGER_OPTS="${HADOOP_JVM_SECURITY_OPTS} ${HADOOP_JVM_GC_OPT}"
export YARN_NODEMANAGER_OPTS="${HADOOP_JVM_SECURITY_OPTS} ${HADOOP_JVM_GC_OPT}"
export YARN_PROXYSERVER_OPTS="${HADOOP_JVM_SECURITY_OPTS} ${HADOOP_JVM_GC_OPT}"
export HADOOP_JOB_HISTORYSERVER_OPTS="${HADOOP_JVM_SECURITY_OPTS} ${HADOOP_JVM_GC_OPT}"
export HADOOP_HEAPSIZE="1536"
export YARN_HEAPSIZE="1024"
export HBASE_OPTS="-XX:MaxDirectMemorySize=128m ${HADOOP_JVM_SECURITY_OPTS} ${HADOOP_JVM_GC_OPT}"
export HBASE_HEAPSIZE="2048"
export HBASE_MANAGES_ZK=false
#export YARN_RESOURCEMANAGER_HEAPSIZE=
#export YARN_NODEMANAGER_HEAPSIZE=
#export YARN_PROXYSERVER_HEAPSIZE=
#export HADOOP_JOB_HISTORYSERVER_HEAPSIZE=

 

(6) 安装上述“编译native lib”步骤中所安装过的包

yum install snappy
yum install snappy-devel
yum install glibc-headers
yum install gcc-c++
yum install cmake
yum install openssl
yum install openssl-devel
yum install zlib
yum install zlib-devel

至此基本的环境准备好了。接下来是具体的配置

 

4. 进入hadoop程序目录,配置hadoop (参考文档:http://hadoop.apache.org/docs/stable/hadoop-project-dist/hadoop-common/ClusterSetup.html)

(1) 修改配置文件

  a. etc/hadoop/core-site.xml 

<configuration>
    <property>  
         <name>fs.defaultFS</name>  
         <value>hdfs://hbase-0:9000/</value>  
    </property>
    <property>
        <name>io.file.buffer.size</name>
        <value>131072</value>
    </property>
    <property>
        <name>hadoop.tmp.dir</name>
        <value>/hbase/hdfstmp</value>
    </property>
    <property>
        <name>hadoop.logfile.size</name>
        <value>104857600</value>
    </property>
    <property>
        <name>hadoop.logfile.count</name>
        <value>20</value>
    </property>
    <property>
        <name>io.bytes.per.checksum</name>
        <value>1024</value>
    </property>
    <property>
        <name>io.skip.checksum.errors</name>
        <value>false</value>
    </property>
    <property>
        <name>io.serializations</name>
        <value>org.apache.hadoop.io.serializer.WritableSerialization</value>
    </property>
    <property>
        <name>io.seqfile.compress.blocksize</name>
        <value>1048576</value>
    </property>
    <property>
        <name>io.compression.codecs</name>
        <value>org.apache.hadoop.io.compress.DefaultCodec,org.apache.hadoop.io.compress.GzipCodec,org.apache.hadoop.io.compress.SnappyCodec</value>
    </property>
</configuration>

 

  b. etc/hadoop/hdfs-site.xml

<configuration>
    <property>
        <name>dfs.replication</name>
        <value>3</value>
    </property>
    <property>
        <name>dfs.blocksize</name>
        <value>2147483648</value>
    </property>
    <property>
        <name>dfs.namenode.handler.count</name>
        <value>128</value>
    </property>
    
    <!-- config for NameNode below -->
    <property>
        <name>dfs.namenode.name.dir</name>
        <value>/hbase/hdfsmeta</value>
    </property>

    <!-- config for DataNode below -->
    <property>
        <name>dfs.datanode.data.dir</name>
        <value>/hbase/hdfs</value>
    </property>
</configuration>

 

  c. etc/hadoop/yarn-site.xml

<configuration>
<!-- Site specific YARN configuration properties -->
    <!-- config for Node Manager below -->
    <property>
        <name>yarn.nodemanager.aux-services</name>
        <value>mapreduce_shuffle</value>
    </property>
    <property>
        <name>yarn.nodemanager.aux-services.mapreduce.shuffle.class</name>
        <value>org.apache.hadoop.mapred.ShuffleHandler</value>
    </property>
    <property>
        <name>yarn.nodemanager.resource.memory-mb</name>
        <value>2048</value>
    </property>
    <property>
        <name>yarn.nodemanager.local-dirs</name>
        <value>/hbase/yarnlocal</value>
    </property>
    <property>
        <name>yarn.nodemanager.log-dirs</name>
        <value>/hbase/yarnlog</value>
    </property>
    <property>
        <name>yarn.nodemanager.log.retain-seconds</name>
        <value>10800</value>
    </property>
    
    <!-- config for Resource Manager below -->
    <property>
        <name>yarn.resourcemanager.address</name>
        <value>hbase-r:18040</value>
    </property>
    <property>
        <name>yarn.resourcemanager.scheduler.address</name>
        <value>hbase-r:18030</value>
    </property>
    <property>
        <name>yarn.resourcemanager.resource-tracker.address</name>
        <value>hbase-r:18025</value>
    </property>
    <property>
        <name>yarn.resourcemanager.admin.address</name>
        <value>hbase-r:18141</value>
    </property>
    <property>
        <name>yarn.resourcemanager.webapp.address</name>
        <value>hbase-r:18088</value>
    </property>

    <property>
        <name>yarn.resourcemanager.scheduler.class</name>
        <value>org.apache.hadoop.yarn.server.resourcemanager.scheduler.capacity.CapacityScheduler</value>
    </property>
    <property>
        <name>yarn.scheduler.minimum-allocation-mb</name>
        <value>256</value>
    </property>
    <property>
        <name>yarn.scheduler.maximum-allocation-mb</name>
        <value>2048</value>
    </property>
    
    
</configuration>

 

  d. etc/hadoop/mapred-site.xml

<configuration>
    <!-- config for MapReduce Applications below -->
    <property>
        <name>mapreduce.framework.name</name>
        <value>yarn</value>
    </property>
    <property>
      <name>mapred.output.compress</name>
      <value>true</value>
      <description>Should the job outputs be compressed?</description>
    </property>
    <property>
        <name>mapred.output.compression.type</name>
        <value>RECORD</value>
    </property>
    <property>
        <name>mapred.output.compression.codec</name>
        <value>org.apache.hadoop.io.compress.SnappyCodec</value>
    </property>
    <property>
        <name>mapred.compress.map.output</name>
        <value>true</value>
    </property>
    <property>
        <name>mapred.map.output.compression.codec</name>
        <value>org.apache.hadoop.io.compress.SnappyCodec</value>
    </property>
    <property>
        <name>mapreduce.map.memory.mb</name>
        <value>1536</value>
    </property>
    <property>
        <name>mapreduce.map.java.opts</name>
        <value>-Xmx1024M</value>
    </property>
    <property>
        <name>mapreduce.reduce.memory.mb</name>
        <value>2048</value>
    </property>
    <property>
        <name>mapreduce.reduce.java.opts</name>
        <value>-Xmx2048M</value>
    </property>
    <property>
        <name>mapreduce.task.io.sort.mb</name>
        <value>512</value>
    </property>
    <property>
        <name>mapreduce.task.io.sort.factor</name>
        <value>100</value>
    </property>
    <property>
        <name>mapreduce.reduce.shuffle.parallelcopies</name>
        <value>50</value>
    </property>
    
    <!-- config for MapReduce JobHistory Server below -->
    <property>
        <name>mapreduce.jobhistory.intermediate-done-dir</name>
        <value>/hbase/mr-history/tmp</value>
    </property>
    <property>
        <name>mapreduce.jobhistory.done-dir</name>
        <value>/hbase/mr-history/done</value>
    </property>

 

  f. etc/hadoop/slaves

  将作为DataNode的host name添加进去:

hbase-1
hbase-2
hbase-3

 

  

  g. 上述配置好的hadoop目录复制到各台服务器。

5. 进入zookeeper目录,修改配置文件

  (1) cp conf/zoo_sample.cfg conf/zoo.cfg,修改conf/zoo.cfg以下配置:

dataDir=/hbase/zookeeperdata
dataLogDir=/hbase/zookeeperlog

server.1=hbase-0:2888:3888
server.2=hbase-r:2888:3888
server.3=hbase-1:2888:3888
server.4=hbase-2:2888:3888
server.5=hbase-3:2888:3888

  (2) 在各台机器上 /hbase/zookeeperdata/下建立文件名myid,分别写入对应的1,2,3,4,5 ZooKeeper序号。

 

6. 进入hbase程序目录,配置hbase (参考文档:http://hbase.apache.org/book.html#quickstart-fully-distributed )

  本次配置的架构为:hbase-0为Master,hbase-1、hbase-2、 hbase-3为RegionServer,ZooKeeper配置在所有服务器(文档中推荐3,5,7)。

  (1) 修改配置文件

  a. conf/regionservers

hbase-1
hbase-2
hbase-3

  b. conf/backup-masters

  将hbase-r设置为master的backup

hbase-r

 

  c. conf/hbase-env.sh

export HBASE_OPTS="-XX:+UseConcMarkSweepGC ${HBASE_OPTS}"

 

  d. conf/hbase-site.xml

 

<configuration>
    <property>
        <name>hbase.tmp.dir</name>
        <value>/hbase/hbasetmp</value>
    </property>
    <property>
      <name>hbase.rootdir</name>
      <value>hdfs://hbase-0:9000/hbase</value>
    </property>
    <property>
        <name>hbase.cluster.distributed</name>
        <value>true</value>
    </property>
    <property>
        <name>hbase.master.info.port</name>
        <value>16010</value>
    </property>
    <property>
        <name>hbase.master.info.bindAddress</name>
        <value>0.0.0.0</value>
    </property>
    <property>
        <name>hbase.column.max.version</name>
        <value>1</value>
    </property>
    
    <!-- for Zoo Keeper -->
    <property>
        <name>hbase.zookeeper.quorum</name>
        <value>hbase-0,hbase-r,hbase-1,hbase-2,hbase-3</value>
    </property>
    <property>
      <name>hbase.zookeeper.property.dataDir</name>
      <value>/hbase/zookeeperdata</value>
    </property>
    <property>
        <name>hbase.zookeeper.property.clientPort</name>
        <value>2181</value>
    </property>
    <property>
        <name>hbase.zookeeper.property.maxClientCnxns</name>
        <value>300</value>
    </property>
    <property>
        <name>zookeeper.session.timeout</name>
        <value>90000</value>
    </property>
    
    <!-- setting for region server -->
    <property>
        <name>hbase.hregion.max.filesize</name>
        <value>2147483648</value>
    </property>
    <property>
        <name>hbase.regionserver.port</name>
        <value>16020</value>
    </property>
    <property>
        <name>hbase.regionserver.codecs</name>
        <value>snappy,lz4</value>
    </property>
    <property>
        <name>hbase.regionserver.info.port</name>
        <value>16030</value>
    </property>
    <property>
        <name>hbase.regionserver.info.bindAddress</name>
        <value>hbase-me</value>
    </property>
    <property>    
        <name>hbase.regionserver.handler.count</name>    
        <value>128</value>    
    </property>
    <property>
        <name>hbase.ipc.server.callqueue.read.ratio</name>
        <value>0.2</value>
    </property>
    <property>
        <name>hbase.ipc.server.callqueue.scan.ratio</name>
        <value>0.2</value>
    </property>
    <property>
        <name>hbase.zookeeper.peerport</name>
        <value>2888</value>
    </property>
    <property>
        <name>hbase.zookeeper.leaderport</name>
        <value>3888</value>
    </property>
    <property>
        <name>hbase.offpeak.start.hour</name>
        <value>0</value>
    </property>
    <property>
        <name>hbase.offpeak.end.hour</name>
        <value>7</value>
    </property>
    <property>
        <name>hbase.rpc.timeout</name>
        <value>60000</value>
    </property>
    <property>
        <name>hbase.rpc.shortoperation.timeout</name>
        <value>10000</value>
    </property>
    <property>
        <name>hbase.ipc.client.tcpnodelay</name>
        <value>true</value>
    </property>
    <property>
        <name>hbase.rest.port</name>
        <value>19080</value>
    </property>
    
    <!-- setting for client -->
    <property>
      <name>hbase.client.pause</name>
      <value>100</value>
    </property>
    <property>
      <name>hbase.ipc.client.tcpnodelay</name>
      <value>false</value>
    </property> 
    <property>
      <name>ipc.ping.interval</name>
      <value>10000</value>
    </property>    
    <property>
        <name>hbase.client.write.buffer</name>
        <value>2097152</value>
    </property>
    <property>
        <name>hbase.client.retries.number</name>
        <value>35</value>
    </property>
    <property>
        <name>hbase.client.max.total.tasks</name>
        <value>100</value>
    </property>
    <property>
        <name>hbase.client.max.perserver.tasks</name>
        <value>20</value>
    </property>
    <property>
        <name>hbase.client.max.perregion.tasks</name>
        <value>1</value>
    </property>
    <property>
        <name>hbase.client.scanner.caching</name>
        <value>100</value>
    </property>
        
</configuration>

 

7. 检查某些系统设置

 (1) 各台服务器关闭防火墙,或者将下列端口加入防火墙允许访问列表(/etc/sysconfig/iptables)

  建议在防火墙上设置所有机器的内网IP,可以互相访问所有端口(单独设置特定的端口范围太过麻烦了,hadoop, hbase, spark, yarn, zookeeper等各种监听端口太多了)。

 以下列表中部分端口号没有在配置的xml出现过,是hadoop或hbase的缺省值(不能保证已经列出了全部监听端口号,仅供参考)。

2181 hbase.zookeeper.property.clientPort
2888 hbase.zookeeper.peerport
3888 hbase.zookeeper.leaderport
9000 fs.defaultFS
10020 mapreduce.jobhistory.address
16010 hbase.master.info.port
16020 hbase.regionserver.port
16030 hbase.regionserver.info.port
18025 yarn.resourcemanager.resource-tracker.address
18030 yarn.resourcemanager.scheduler.address
18040 yarn.resourcemanager.address
18088 yarn.resourcemanager.webapp.address
18141 yarn.resourcemanager.admin.address
19080 hbase.rest.port
19888 mapreduce.jobhistory.webapp.address
50010 hdfs client
60000 hbase.master.port
60020 hbase.regionserver.port
33651 catalog.CatalogTracker
40641 zookeeper.ClientCnxn

   (2) 检查ulimit, /etc/security/limits.conf

   此处暂时设置为(假设启动hadoop的用户为admin)

admin        -       nofile  65536
admin        -       nproc   65536

   

 8. 启动hadoop

  (1) 初次启动之前,在hbase-0上格式化NameNode

$HADOOP_PREFIX/bin/hdfs namenode -format hbase-0

 

  在hbase-r上格式化NameNode

$HADOOP_PREFIX/bin/hdfs namenode -format hbase-r

 

  (2) 在hbase-0启动NameNode

$HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs start namenode

  (3) 分别在hbase-1,hbase-2, hbase-3 上启动DataNode

$HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs start datanode

  (4) 在hbase-r上启动ResourceManager

$HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR start resourcemanager

  (5) 在hbase-1, hbase-2, hbase-3上启动NodeManager

$HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR start nodemanager

  (6) 在hbase-r上启动Yarn的WebAppProxy Server

$HADOOP_YARN_HOME/sbin/yarn-daemon.sh start proxyserver --config $HADOOP_CONF_DIR

  (7) 在hbase-r上启动MapReduce JobHistory Server

$HADOOP_PREFIX/sbin/mr-jobhistory-daemon.sh start historyserver --config $HADOOP_CONF_DIR

 

9. 停止hadoop

  (1) hbase-0上停止NameNode

$HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs stop namenode

  (2) 分别在hbase-1,hbase-2, hbase-3 上停止DataNode

$HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs stop datanode

  (3) 在hbase-r上停止ResourceManager

$HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR stop resourcemanager

  (4) 在hbase-1, hbase-2, hbase-3上停止NodeManager

$HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR stop nodemanager

  (5) 在hbase-r上停止Yarn的WebAppProxy Server

$HADOOP_YARN_HOME/sbin/yarn-daemon.sh stop proxyserver --config $HADOOP_CONF_DIR

  (6) 在hbase-r上停止MapReduce JobHistory Server

$HADOOP_PREFIX/sbin/mr-jobhistory-daemon.sh stop historyserver --config $HADOOP_CONF_DIR

 

10. 启动zookeeper

  (1) 在所有机器上启动zookeeper

$ZOOKEEPER_HOME/bin/zkServer.sh start

 

  (2) 在各台机器上确认启动状态

$ZOOKEEPER_HOME/bin/zkServer.sh status
JMX enabled by default
Using config: /hbase/zookeeper/bin/../conf/zoo.cfg
Mode: leader

  其中Mode:leader应该只有1个,其他的都是Mode: follower

 

11. 启动hbase

  (1) 在hbase-0上启动hbase。按照文档中的描述,会按照ZooKeeper、Master、RegionServers、Backup masters的顺序启动。

$HBASE_HOME/bin/start-hbase.sh

  (2) 在各个结点上执行 jps指令,确认启动状况

  hbase-0:

$ jps
14761 NameNode
15644 Bootstrap
10145 QuorumPeerMain
21312 HMaster
24155 Jps
16377 Bootstrap

 

  hbase-r:

$ jps
14229 ResourceManager
19301 Bootstrap
1119 Bootstrap
15188 JobHistoryServer
20920 Jps
10831 QuorumPeerMain

 

  hbase-1:

 

$ jps
9015 HRegionServer
4402 DataNode
10906 Bootstrap
1145 QuorumPeerMain
15282 Bootstrap
10763 Jps
10876 NodeManager

 

  hbase-2:

$ jps
17856 HRegionServer
13396 DataNode
17399 QuorumPeerMain
18258 Jps
28873 NodeManager

 hbase-3:

 

$ jps
5660 Jps
3825 HRegionServer
24410 Bootstrap
5509 NodeManager
31687 DataNode
28529 QuorumPeerMain
9915 Bootstrap

 

 

可以看到,

hadoop:1个NameNode, 3个DataNode,3个NodeManager,1个ResourceManager, 1个JobHistoryServer

hbase:1个HMaster,3个RegionServer,

5个ZooKeeper,

全部启动完成了。

 

  (2) 停止hbase 

$HBASE_HOME/bin/stop-hbase.sh

 12. 一些基本的测试

  (1) Snappy压缩测试

$ $HBASE_HOME/bin/hbase shell
> create 'test1', { NAME => 'cf1', COMPRESSION => 'SNAPPY' }
> put 'test1', 'row1', 'cf1:a', 'value1'
> scan 'test1'
ROW                      COLUMN+CELL                                                         
row1                    column=cf1:a, timestamp=1414480111169, value=value1                 
1 row(s) in 0.0370 seconds

  (2) 去hadoop看刚才创建的记录

$HADOOP_PREFIX/bin/hdfs dfs -ls /hbase/data/default/test1/
drwxr-xr-x   - admin supergroup          0 2014-10-28 15:07 /hbase/data/default/test1/.tabledesc
drwxr-xr-x   - admin supergroup          0 2014-10-28 15:07 /hbase/data/default/test1/.tmp
drwxr-xr-x   - admin supergroup          0 2014-10-28 15:07 /hbase/data/default/test1/003d38c6f213f8c589eea5065785ce57

 

  (3) 浏览器访问hbase-0:50070/,通过管理界面可以看到Live Nodes: 3

  (4) 浏览器访问hbase-0:16010/,通过管理界面可以看到Region Servers 一共3台: hbase-1, hbase-2, hbase-3全部运行中。

 

13. 为了方便,把启动和停止做成了2个脚本(启动所有的hadoop, zookeeper, hbase)。

  (1) 启动

ssh hbase-0 $HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs start namenode
ssh hbase-1 $HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs start datanode
ssh hbase-2 $HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs start datanode
ssh hbase-3 $HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs start datanode
ssh hbase-r $HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR start resourcemanager
ssh hbase-1 $HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR start nodemanager
ssh hbase-2 $HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR start nodemanager
ssh hbase-3 $HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR start nodemanager
ssh hbase-r $HADOOP_YARN_HOME/sbin/yarn-daemon.sh start proxyserver --config $HADOOP_CONF_DIR
ssh hbase-r $HADOOP_PREFIX/sbin/mr-jobhistory-daemon.sh start historyserver --config $HADOOP_CONF_DIR
ssh hbase-0 $ZOOKEEPER_HOME/bin/zkServer.sh start
ssh hbase-r $ZOOKEEPER_HOME/bin/zkServer.sh start
ssh hbase-1 $ZOOKEEPER_HOME/bin/zkServer.sh start
ssh hbase-2 $ZOOKEEPER_HOME/bin/zkServer.sh start
ssh hbase-3 $ZOOKEEPER_HOME/bin/zkServer.sh start
ssh hbase-0 $HBASE_HOME/bin/start-hbase.sh

 

  (2) 停止

ssh hbase-0 $HBASE_HOME/bin/stop-hbase.sh
ssh hbase-0 $ZOOKEEPER_HOME/bin/zkServer.sh stop
ssh hbase-r $ZOOKEEPER_HOME/bin/zkServer.sh stop
ssh hbase-1 $ZOOKEEPER_HOME/bin/zkServer.sh stop
ssh hbase-2 $ZOOKEEPER_HOME/bin/zkServer.sh stop
ssh hbase-3 $ZOOKEEPER_HOME/bin/zkServer.sh stop
ssh hbase-0 $HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs stop namenode
ssh hbase-1 $HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs stop datanode
ssh hbase-2 $HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs stop datanode
ssh hbase-3 $HADOOP_PREFIX/sbin/hadoop-daemon.sh --config $HADOOP_CONF_DIR --script hdfs stop datanode
ssh hbase-r $HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR stop resourcemanager
ssh hbase-1 $HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR stop nodemanager
ssh hbase-2 $HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR stop nodemanager
ssh hbase-3 $HADOOP_YARN_HOME/sbin/yarn-daemon.sh --config $HADOOP_CONF_DIR stop nodemanager
ssh hbase-r $HADOOP_YARN_HOME/sbin/yarn-daemon.sh stop proxyserver --config $HADOOP_CONF_DIR
ssh hbase-r $HADOOP_PREFIX/sbin/mr-jobhistory-daemon.sh stop historyserver --config $HADOOP_CONF_DIR

 

 

-------------------------------------- 结束 -----------------------------------------

此次部署过程遇到了各种各样的错误,通过查看启动日志,搜索其他人的博文,最终成功安装上去了。希望本文能为后来者省点时间,也为下次自己再次部署的时候作为参考。

 

posted @ 2014-10-28 15:55  Beef Liu  阅读(848)  评论(0编辑  收藏  举报