共计 23467 个字符，预计需要花费 59 分钟才能阅读完成。
1. Hadoop 的 HA 机制
前言：正式引入 HA 机制是从 hadoop2.0 开始，之前的版本中没有 HA 机制
1.1. HA 的运作机制
（1）hadoop-HA 集群运作机制介绍
所谓 HA，即高可用（7*24 小时不中断服务）
实现高可用最关键的是消除单点故障
hadoop-ha 严格来说应该分成各个组件的 HA 机制——HDFS 的 HA、YARN 的 HA
（2）HDFS 的 HA 机制详解
通过双 namenode 消除单点故障
双 namenode 协调工作的要点：
A、元数据管理方式需要改变：
内存中各自保存一份元数据
Edits 日志只能有一份，只有 Active 状态的 namenode 节点可以做写操作
两个 namenode 都可以读取 edits
共享的 edits 放在一个共享存储中管理（qjournal 和 NFS 两个主流实现）
B、需要一个状态管理功能模块
实现了一个 zkfailover，常驻在每一个 namenode 所在的节点
每一个 zkfailover 负责监控自己所在 namenode 节点，利用 zk 进行状态标识
当需要进行状态切换时，由 zkfailover 来负责切换
切换时需要防止 brain split 现象的发生
1.2. HDFS-HA 图解
2. 主机规划
主机名称 | 外网 IP | 内网 IP | 操作系统 | 备注 | 安装软件 | 运行进程 |
mini01 | 10.0.0.111 | 172.16.1.111 | CentOS 7.4 | ssh port:22 | jdk、hadoop | NameNode、DFSZKFailoverController(zkfc) |
mini02 | 10.0.0.112 | 172.16.1.112 | CentOS 7.4 | ssh port:22 | jdk、hadoop | NameNode、DFSZKFailoverController(zkfc) |
mini03 | 10.0.0.113 | 172.16.1.113 | CentOS 7.4 | ssh port:22 | jdk、hadoop、zookeeper | ResourceManager |
mini04 | 10.0.0.114 | 172.16.1.114 | CentOS 7.4 | ssh port:22 | jdk、hadoop、zookeeper | ResourceManager |
mini05 | 10.0.0.115 | 172.16.1.115 | CentOS 7.4 | ssh port:22 | jdk、hadoop、zookeeper | DataNode、NodeManager、JournalNode、QuorumPeerMain |
mini06 | 10.0.0.116 | 172.16.1.116 | CentOS 7.4 | ssh port:22 | jdk、hadoop、zookeeper | DataNode、NodeManager、JournalNode、QuorumPeerMain |
mini07 | 10.0.0.117 | 172.16.1.117 | CentOS 7.4 | ssh port:22 | jdk、hadoop、zookeeper | DataNode、NodeManager、JournalNode、QuorumPeerMain |
注意：针对 HA 模式，就不需要 SecondaryNameNode 了，因为 STANDBY 状态的 namenode 会负责做 checkpoint
Linux 添加 hosts 信息，保证每台都可以相互 ping 通
[root@mini01 ~]# cat /etc/hosts
127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
10.0.0.111 mini01
10.0.0.112 mini02
10.0.0.113 mini03
10.0.0.114 mini04
10.0.0.115 mini05
10.0.0.116 mini06
10.0.0.117 mini07
Windows 的 hosts 文件修改
# 文件位置 C:\Windows\System32\drivers\etc 在 hosts 中追加如下内容
…………………………………………
10.0.0.111 mini01
10.0.0.112 mini02
10.0.0.113 mini03
10.0.0.114 mini04
10.0.0.115 mini05
10.0.0.116 mini06
10.0.0.117 mini07
3. 添加用户账号
# 使用一个专门的用户，避免直接使用 root 用户
# 添加用户、指定家目录并指定用户密码
useradd -d /app yun && echo ‘123456’ | /usr/bin/passwd –stdin yun
# sudo 提权
echo “yun ALL=(ALL) NOPASSWD: ALL” >> /etc/sudoers
# 让其它普通用户可以进入该目录查看信息
chmod 755 /app/
4. 实现 yun 用户免秘钥登录
要求：根据规划实现 mini01 到 mini01、mini02、mini03、mini04、mini05、mini06、mini07 免秘钥登录
实现 mini02 到 mini01、mini02、mini03、mini04、mini05、mini06、mini07 免秘钥登录
实现 mini03 到 mini01、mini02、mini03、mini04、mini05、mini06、mini07 免秘钥登录
实现 mini04 到 mini01、mini02、mini03、mini04、mini05、mini06、mini07 免秘钥登录
实现 mini05 到 mini01、mini02、mini03、mini04、mini05、mini06、mini07 免秘钥登录
实现 mini06 到 mini01、mini02、mini03、mini04、mini05、mini06、mini07 免秘钥登录
实现 mini07 到 mini01、mini02、mini03、mini04、mini05、mini06、mini07 免秘钥登录
# 可以使用 ip 也可以是 hostname 但是由于我们计划使用的是 hostname 方式交互，所以使用 hostname
# 同时 hostname 方式分发，可以通过 hostname 远程登录，也可以 IP 远程登录
具体过程就不多说了，请参见 https://www.linuxidc.com/Linux/2018-08/153353.htm
5. Jdk【java8】
具体过程就不多说了，请参见 https://www.linuxidc.com/Linux/2018-08/153353.htm
6. Zookeeper 部署
根据规划 zookeeper 部署在 mini03、mini04、mini05、mini06、mini07 上
6.1. 配置信息
[yun@mini03 conf]$ pwd
/app/zookeeper/conf
[yun@mini03 conf]$ vim zoo.cfg
# 单个客户端与单台服务器之间的连接数的限制，是 ip 级别的，默认是 60，如果设置为 0，那么表明不作任何限制。
maxClientCnxns=1500
# The number of milliseconds of each tick
tickTime=2000
# The number of ticks that the initial
# synchronization phase can take
initLimit=10
# The number of ticks that can pass between
# sending a request and getting an acknowledgement
syncLimit=5
# the directory where the snapshot is stored.
# do not use /tmp for storage, /tmp here is just
# example sakes.
# dataDir=/tmp/zookeeper
dataDir=/app/bigdata/zookeeper/data
# the port at which the clients will connect
clientPort=2181
#
# Be sure to read the maintenance section of the
# administrator guide before turning on autopurge.
#
# http://zookeeper.apache.org/doc/current/zookeeperAdmin.html#sc_maintenance
#
# The number of snapshots to retain in dataDir
#autopurge.snapRetainCount=3
# Purge task interval in hours
# Set to “0” to disable auto purge feature
#autopurge.purgeInterval=1
# leader 和 follow 通信端口和投票选举端口
server.3=mini03:2888:3888
server.4=mini04:2888:3888
server.5=mini05:2888:3888
server.6=mini06:2888:3888
server.7=mini07:2888:3888
6.2. 添加 myid 文件
[yun@mini03 data]$ pwd
/app/bigdata/zookeeper/data
[yun@mini03 data]$ vim myid # 其中 mini03 的 myid 为 3；mini04 的 myid 为 4；mini05 的 myid 为 5；mini06 的 myid 为 6；mini07 的 myid 为 7
3
6.3. 启动 zk 服务
# 依次在启动 mini03、mini04、mini05、mini06、mini07 zk 服务
[yun@mini03 ~]$ cd zookeeper/bin/
[yun@mini03 bin]$ pwd
/app/zookeeper/bin
[yun@mini03 bin]$ ll
total 56
-rwxr-xr-x 1 yun yun 238 Oct 1 2012 README.txt
-rwxr-xr-x 1 yun yun 1909 Oct 1 2012 zkCleanup.sh
-rwxr-xr-x 1 yun yun 1049 Oct 1 2012 zkCli.cmd
-rwxr-xr-x 1 yun yun 1512 Oct 1 2012 zkCli.sh
-rwxr-xr-x 1 yun yun 1333 Oct 1 2012 zkEnv.cmd
-rwxr-xr-x 1 yun yun 2599 Oct 1 2012 zkEnv.sh
-rwxr-xr-x 1 yun yun 1084 Oct 1 2012 zkServer.cmd
-rwxr-xr-x 1 yun yun 5467 Oct 1 2012 zkServer.sh
-rw-rw-r– 1 yun yun 17522 Jun 28 21:01 zookeeper.out
[yun@mini03 bin]$ ./zkServer.sh start
JMX enabled by default
Using config: /app/zookeeper/bin/../conf/zoo.cfg
Starting zookeeper … STARTED
6.4. 查询运行状态
# 其中 mini03、mini04、mini06、mini07 状态如下
[yun@mini03 bin]$ ./zkServer.sh status
JMX enabled by default
Using config: /app/zookeeper/bin/../conf/zoo.cfg
Mode: follower
# 其中 mini05 状态如下
[yun@mini05 bin]$ ./zkServer.sh status
JMX enabled by default
Using config: /app/zookeeper/bin/../conf/zoo.cfg
Mode: leader
PS：4 个 follower 1 个 leader
7. Hadoop 部署与配置修改
注意：每台机器的 Hadoop 以及配置相同
7.1. 部署
[yun@mini01 software]$ pwd
/app/software
[yun@mini01 software]$ ll
total 194152
-rw-r–r– 1 yun yun 198811365 Jun 8 16:36 CentOS-7.4_hadoop-2.7.6.tar.gz
[yun@mini01 software]$ tar xf CentOS-7.4_hadoop-2.7.6.tar.gz
[yun@mini01 software]$ mv hadoop-2.7.6/ /app/
[yun@mini01 software]$ cd
[yun@mini01 ~]$ ln -s hadoop-2.7.6/ hadoop
[yun@mini01 ~]$ ll
total 4
lrwxrwxrwx 1 yun yun 13 Jun 9 16:21 hadoop -> hadoop-2.7.6/
drwxr-xr-x 9 yun yun 149 Jun 8 16:36 hadoop-2.7.6
lrwxrwxrwx 1 yun yun 12 May 26 11:18 jdk -> jdk1.8.0_112
drwxr-xr-x 8 yun yun 255 Sep 23 2016 jdk1.8.0_112
7.2. 环境变量
[root@mini01 profile.d]# pwd
/etc/profile.d
[root@mini01 profile.d]# vim hadoop.sh
export HADOOP_HOME=”/app/hadoop”
export PATH=$HADOOP_HOME/bin:$HADOOP_HOME/sbin:$PATH
[root@mini01 profile.d]# source /etc/profile # 生效
7.3. core-site.xml
[yun@mini01 hadoop]$ pwd
/app/hadoop/etc/hadoop
[yun@mini01 hadoop]$ vim core-site.xml
<?xml version=”1.0″ encoding=”UTF-8″?>
<?xml-stylesheet type=”text/xsl” href=”https://www.linuxidc.com/Linux/2018-08/configuration.xsl”?>
<!–
……………………
–>
<!– Put site-specific property overrides in this file. –>
<configuration>
<!– 指定 hdfs 的 nameservice 为 bi –>
<property>
<name>fs.defaultFS</name>
<value>hdfs://bi/</value>
</property>
<!– 指定 hadoop 临时目录 –>
<property>
<name>hadoop.tmp.dir</name>
<value>/app/hadoop/tmp</value>
</property>
<!– 指定 zookeeper 地址 –>
<property>
<name>ha.zookeeper.quorum</name>
<value>mini03:2181,mini04:2181,mini05:2181,mini06:2181,mini07:2181</value>
</property>
</configuration>
7.4. hdfs-site.xml
[yun@mini01 hadoop]$ pwd
/app/hadoop/etc/hadoop
[yun@mini01 hadoop]$ vim hdfs-site.xml
<?xml version=”1.0″ encoding=”UTF-8″?>
<?xml-stylesheet type=”text/xsl” href=”https://www.linuxidc.com/Linux/2018-08/configuration.xsl”?>
<!–
……………………
–>
<!– Put site-specific property overrides in this file. –>
<configuration>
<!– 指定 hdfs 的 nameservice 为 bi，需要和 core-site.xml 中的保持一致 –>
<property>
<name>dfs.nameservices</name>
<value>bi</value>
</property>
<!– bi 下面有两个 NameNode，分别是 nn1，nn2 –>
<property>
<name>dfs.ha.namenodes.bi</name>
<value>nn1,nn2</value>
</property>
<!– nn1 的 RPC 通信地址 –>
<property>
<name>dfs.namenode.rpc-address.bi.nn1</name>
<value>mini01:9000</value>
</property>
<!– nn1 的 http 通信地址 –>
<property>
<name>dfs.namenode.http-address.bi.nn1</name>
<value>mini01:50070</value>
</property>
<!– nn2 的 RPC 通信地址 –>
<property>
<name>dfs.namenode.rpc-address.bi.nn2</name>
<value>mini02:9000</value>
</property>
<!– nn2 的 http 通信地址 –>
<property>
<name>dfs.namenode.http-address.bi.nn2</name>
<value>mini02:50070</value>
</property>
<!– 指定 NameNode 的 edits 元数据在 JournalNode 上的存放位置 –>
<property>
<name>dfs.namenode.shared.edits.dir</name>
<value>qjournal://mini05:8485;mini06:8485;mini07:8485/bi</value>
</property>
<!– 指定 JournalNode 在本地磁盘存放数据的位置 –>
<property>
<name>dfs.journalnode.edits.dir</name>
<value>/app/hadoop/journaldata</value>
</property>
<!– 开启 NameNode 失败自动切换 –>
<property>
<name>dfs.ha.automatic-failover.enabled</name>
<value>true</value>
</property>
<!– 配置失败自动切换实现方式 –>
<property>
<name>dfs.client.failover.proxy.provider.bi</name>
<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
</property>
<!– 配置隔离机制方法，多个机制用换行分割，即每个机制暂用一行 –>
<!– 其中 shell(/bin/true) 表示可执行一个脚本 比如 shell(/app/yunwei/hadoop_fence.sh) –>
<property>
<name>dfs.ha.fencing.methods</name>
<value>
sshfence
shell(/bin/true)
</value>
</property>
<!– 使用 sshfence 隔离机制时需要 ssh 免登陆 –>
<property>
<name>dfs.ha.fencing.ssh.private-key-files</name>
<value>/app/.ssh/id_rsa</value>
</property>
<!– 配置 sshfence 隔离机制超时时间 单位：毫秒 –>
<property>
<name>dfs.ha.fencing.ssh.connect-timeout</name>
<value>30000</value>
</property>
</configuration>
7.5. mapred-site.xml
[yun@mini01 hadoop]$ pwd
/app/hadoop/etc/hadoop
[yun@mini01 hadoop]$ cp -a mapred-site.xml.template mapred-site.xml
[yun@mini01 hadoop]$ vim mapred-site.xml
<?xml version=”1.0″?>
<?xml-stylesheet type=”text/xsl” href=”https://www.linuxidc.com/Linux/2018-08/configuration.xsl”?>
<!–
……………………
–>
<!– Put site-specific property overrides in this file. –>
<configuration>
<!– 指定 mr 框架为 yarn 方式 –>
<property>
<name>mapreduce.framework.name</name>
<value>yarn</value>
</property>
</configuration>
7.6. yarn-site.xml
[yun@mini01 hadoop]$ pwd
/app/hadoop/etc/hadoop
[yun@mini01 hadoop]$ vim yarn-site.xml
<?xml version=”1.0″?>
<!–
……………………
–>
<configuration>
<!– Site specific YARN configuration properties –>
<!– 开启 RM 高可用 –>
<property>
<name>yarn.resourcemanager.ha.enabled</name>
<value>true</value>
</property>
<!– 指定 RM 的 cluster id –>
<property>
<name>yarn.resourcemanager.cluster-id</name>
<value>yrc</value>
</property>
<!– 指定 RM 的名字 –>
<property>
<name>yarn.resourcemanager.ha.rm-ids</name>
<value>rm1,rm2</value>
</property>
<!– 分别指定 RM 的地址 –>
<property>
<name>yarn.resourcemanager.hostname.rm1</name>
<value>mini03</value>
</property>
<property>
<name>yarn.resourcemanager.hostname.rm2</name>
<value>mini04</value>
</property>
<!– 指定 zk 集群地址 –>
<property>
<name>yarn.resourcemanager.zk-address</name>
<value>mini03:2181,mini04:2181,mini05:2181,mini06:2181,mini07:2181</value>
</property>
<!– reduce 获取数据的方式 –>
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
</configuration>
7.7. 修改 slaves
slaves 是指定子节点的位置，因为要在 mini01 上启动 HDFS、在 mini03 启动 yarn，所以 mini01 上的 slaves 文件指定的是 datanode 的位置，mini03 上的 slaves 文件指定的是 nodemanager 的位置
[yun@mini01 hadoop]$ pwd
/app/hadoop/etc/hadoop
[yun@mini01 hadoop]$ vim slaves
mini05
mini06
mini07
PS：改后配置后，将这些配置拷到其他 Hadoop 机器
8. 启动相关服务
注意：第一次启动时严格按照下面的步骤！！！！！！！
8.1. 启动 zookeeper 集群
前面已经启动了，这里就不说了
8.2. 启动 journalnode
# 根据规划在 mini05、mini06、mini07 启动
# 在第一次格式化的时候需要先启动 journalnode 之后就不必了
[yun@mini05 ~]$ hadoop-daemon.sh start journalnode # 已经配置环境变量，所以不用进入到响应的目录
starting journalnode, logging to /app/hadoop-2.7.6/logs/hadoop-yun-journalnode-mini05.out
[yun@mini05 ~]$ jps
1281 QuorumPeerMain
1817 Jps
1759 JournalNode
8.3. 格式化 HDFS
# 在 mini01 上执行命令
[yun@mini01 ~]$ hdfs namenode -format
18/06/30 18:29:12 INFO namenode.NameNode: STARTUP_MSG:
/************************************************************
STARTUP_MSG: Starting NameNode
STARTUP_MSG: host = mini01/10.0.0.111
STARTUP_MSG: args = [-format]
STARTUP_MSG: version = 2.7.6
STARTUP_MSG: classpath = ………………
STARTUP_MSG: build = Unknown -r Unknown; compiled by ‘root’ on 2018-06-08T08:30Z
STARTUP_MSG: java = 1.8.0_112
************************************************************/
18/06/30 18:29:12 INFO namenode.NameNode: registered UNIX signal handlers for [TERM, HUP, INT]
18/06/30 18:29:12 INFO namenode.NameNode: createNameNode [-format]
Formatting using clusterid: CID-2385f26e-72e6-4935-aa09-47848b5ba4be
18/06/30 18:29:13 INFO namenode.FSNamesystem: No KeyProvider found.
18/06/30 18:29:13 INFO namenode.FSNamesystem: fsLock is fair: true
18/06/30 18:29:13 INFO namenode.FSNamesystem: Detailed lock hold time metrics enabled: false
18/06/30 18:29:13 INFO blockmanagement.DatanodeManager: dfs.block.invalidate.limit=1000
18/06/30 18:29:13 INFO blockmanagement.DatanodeManager: dfs.namenode.datanode.registration.ip-hostname-check=true
18/06/30 18:29:13 INFO blockmanagement.BlockManager: dfs.namenode.startup.delay.block.deletion.sec is set to 000:00:00:00.000
18/06/30 18:29:13 INFO blockmanagement.BlockManager: The block deletion will start around 2018 Jun 30 18:29:13
18/06/30 18:29:13 INFO util.GSet: Computing capacity for map BlocksMap
18/06/30 18:29:13 INFO util.GSet: VM type = 64-bit
18/06/30 18:29:13 INFO util.GSet: 2.0% max memory 966.7 MB = 19.3 MB
18/06/30 18:29:13 INFO util.GSet: capacity = 2^21 = 2097152 entries
18/06/30 18:29:13 INFO blockmanagement.BlockManager: dfs.block.access.token.enable=false
18/06/30 18:29:13 INFO blockmanagement.BlockManager: defaultReplication = 3
18/06/30 18:29:13 INFO blockmanagement.BlockManager: maxReplication = 512
18/06/30 18:29:13 INFO blockmanagement.BlockManager: minReplication = 1
18/06/30 18:29:13 INFO blockmanagement.BlockManager: maxReplicationStreams = 2
18/06/30 18:29:13 INFO blockmanagement.BlockManager: replicationRecheckInterval = 3000
18/06/30 18:29:13 INFO blockmanagement.BlockManager: encryptDataTransfer = false
18/06/30 18:29:13 INFO blockmanagement.BlockManager: maxNumBlocksToLog = 1000
18/06/30 18:29:13 INFO namenode.FSNamesystem: fsOwner = yun (auth:SIMPLE)
18/06/30 18:29:13 INFO namenode.FSNamesystem: supergroup = supergroup
18/06/30 18:29:13 INFO namenode.FSNamesystem: isPermissionEnabled = true
18/06/30 18:29:13 INFO namenode.FSNamesystem: Determined nameservice ID: bi
18/06/30 18:29:13 INFO namenode.FSNamesystem: HA Enabled: true
18/06/30 18:29:13 INFO namenode.FSNamesystem: Append Enabled: true
18/06/30 18:29:13 INFO util.GSet: Computing capacity for map INodeMap
18/06/30 18:29:13 INFO util.GSet: VM type = 64-bit
18/06/30 18:29:13 INFO util.GSet: 1.0% max memory 966.7 MB = 9.7 MB
18/06/30 18:29:13 INFO util.GSet: capacity = 2^20 = 1048576 entries
18/06/30 18:29:13 INFO namenode.FSDirectory: ACLs enabled? false
18/06/30 18:29:13 INFO namenode.FSDirectory: XAttrs enabled? true
18/06/30 18:29:13 INFO namenode.FSDirectory: Maximum size of an xattr: 16384
18/06/30 18:29:13 INFO namenode.NameNode: Caching file names occuring more than 10 times
18/06/30 18:29:13 INFO util.GSet: Computing capacity for map cachedBlocks
18/06/30 18:29:13 INFO util.GSet: VM type = 64-bit
18/06/30 18:29:13 INFO util.GSet: 0.25% max memory 966.7 MB = 2.4 MB
18/06/30 18:29:13 INFO util.GSet: capacity = 2^18 = 262144 entries
18/06/30 18:29:13 INFO namenode.FSNamesystem: dfs.namenode.safemode.threshold-pct = 0.9990000128746033
18/06/30 18:29:13 INFO namenode.FSNamesystem: dfs.namenode.safemode.min.datanodes = 0
18/06/30 18:29:13 INFO namenode.FSNamesystem: dfs.namenode.safemode.extension = 30000
18/06/30 18:29:13 INFO metrics.TopMetrics: NNTop conf: dfs.namenode.top.window.num.buckets = 10
18/06/30 18:29:13 INFO metrics.TopMetrics: NNTop conf: dfs.namenode.top.num.users = 10
18/06/30 18:29:13 INFO metrics.TopMetrics: NNTop conf: dfs.namenode.top.windows.minutes = 1,5,25
18/06/30 18:29:13 INFO namenode.FSNamesystem: Retry cache on namenode is enabled
18/06/30 18:29:13 INFO namenode.FSNamesystem: Retry cache will use 0.03 of total heap and retry cache entry expiry time is 600000 millis
18/06/30 18:29:13 INFO util.GSet: Computing capacity for map NameNodeRetryCache
18/06/30 18:29:13 INFO util.GSet: VM type = 64-bit
18/06/30 18:29:13 INFO util.GSet: 0.029999999329447746% max memory 966.7 MB = 297.0 KB
18/06/30 18:29:13 INFO util.GSet: capacity = 2^15 = 32768 entries
18/06/30 18:29:14 INFO namenode.FSImage: Allocated new BlockPoolId: BP-1178102935-10.0.0.111-1530354554626
18/06/30 18:29:14 INFO common.Storage: Storage directory /app/hadoop/tmp/dfs/name has been successfully formatted.
18/06/30 18:29:14 INFO namenode.FSImageFormatProtobuf: Saving image file /app/hadoop/tmp/dfs/name/current/fsimage.ckpt_0000000000000000000 using no compression
18/06/30 18:29:14 INFO namenode.FSImageFormatProtobuf: Image file /app/hadoop/tmp/dfs/name/current/fsimage.ckpt_0000000000000000000 of size 320 bytes saved in 0 seconds.
18/06/30 18:29:15 INFO namenode.NNStorageRetentionManager: Going to retain 1 images with txid >= 0
18/06/30 18:29:15 INFO util.ExitUtil: Exiting with status 0
18/06/30 18:29:15 INFO namenode.NameNode: SHUTDOWN_MSG:
/************************************************************
SHUTDOWN_MSG: Shutting down NameNode at mini01/10.0.0.111
************************************************************/
拷贝到 mini02
# 格式化后会在根据 core-site.xml 中的 hadoop.tmp.dir 配置生成个文件，这里我配置的是 /app/hadoop/tmp，然后将 /app/hadoop/tmp 拷贝到 mini02 的 /app/hadoop/ 下。
# 方法 1：
[yun@mini01 hadoop]$ pwd
/app/hadoop
[yun@mini01 hadoop]$ scp -r tmp/ yun@mini02:/app/hadoop
VERSION 100% 202 189.4KB/s 00:00
seen_txid 100% 2 1.0KB/s 00:00
fsimage_0000000000000000000.md5 100% 62 39.7KB/s 00:00
fsimage_0000000000000000000 100% 320 156.1KB/s 00:00
##########################3
# 方法 2：## 也可以这样，建议 hdfs namenode -bootstrapStandby # 不过需要 mini02 的 Hadoop 起来才行
8.4. 格式化 ZKFC
# 在 mini01 上执行一次即可
[yun@mini01 ~]$ hdfs zkfc -formatZK
18/06/30 18:54:30 INFO tools.DFSZKFailoverController: Failover controller configured for NameNode NameNode at mini01/10.0.0.111:9000
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:zookeeper.version=3.4.6-1569965, built on 02/20/2014 09:09 GMT
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:host.name=mini01
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:java.version=1.8.0_112
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:java.vendor=Oracle Corporation
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:java.home=/app/jdk1.8.0_112/jre
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:java.class.path=……………………
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:java.library.path=/app/hadoop-2.7.6/lib/native
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:java.io.tmpdir=/tmp
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:java.compiler=<NA>
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:os.name=Linux
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:os.arch=amd64
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:os.version=3.10.0-693.el7.x86_64
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:user.name=yun
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:user.home=/app
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Client environment:user.dir=/app/hadoop-2.7.6
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Initiating client connection, connectString=mini03:2181,mini04:2181,mini05:2181,mini06:2181,mini07:2181 sessionTimeout=5000 watcher=org.apache.hadoop.ha.ActiveStandbyElector$WatcherWithClientRef@7f3b84b8
18/06/30 18:54:30 INFO zookeeper.ClientCnxn: Opening socket connection to server mini04/10.0.0.114:2181. Will not attempt to authenticate using SASL (unknown error)
18/06/30 18:54:30 INFO zookeeper.ClientCnxn: Socket connection established to mini04/10.0.0.114:2181, initiating session
18/06/30 18:54:30 INFO zookeeper.ClientCnxn: Session establishment complete on server mini04/10.0.0.114:2181, sessionid = 0x4644fff9cb80000, negotiated timeout = 5000
18/06/30 18:54:30 INFO ha.ActiveStandbyElector: Session connected.
18/06/30 18:54:30 INFO ha.ActiveStandbyElector: Successfully created /hadoop-ha/bi in ZK.
18/06/30 18:54:30 INFO zookeeper.ZooKeeper: Session: 0x4644fff9cb80000 closed
18/06/30 18:54:30 INFO zookeeper.ClientCnxn: EventThread shut down
8.5. 启动 HDFS
# 在 mini01 上执行
[yun@mini01 ~]$ start-dfs.sh
Starting namenodes on [mini01 mini02]
mini01: starting namenode, logging to /app/hadoop-2.7.6/logs/hadoop-yun-namenode-mini01.out
mini02: starting namenode, logging to /app/hadoop-2.7.6/logs/hadoop-yun-namenode-mini02.out
mini07: starting datanode, logging to /app/hadoop-2.7.6/logs/hadoop-yun-datanode-mini07.out
mini06: starting datanode, logging to /app/hadoop-2.7.6/logs/hadoop-yun-datanode-mini06.out
mini05: starting datanode, logging to /app/hadoop-2.7.6/logs/hadoop-yun-datanode-mini05.out
Starting journal nodes [mini05 mini06 mini07]
mini07: journalnode running as process 1691. Stop it first.
mini06: journalnode running as process 1665. Stop it first.
mini05: journalnode running as process 1759. Stop it first.
Starting ZK Failover Controllers on NN hosts [mini01 mini02]
mini01: starting zkfc, logging to /app/hadoop-2.7.6/logs/hadoop-yun-zkfc-mini01.out
mini02: starting zkfc, logging to /app/hadoop-2.7.6/logs/hadoop-yun-zkfc-mini02.out
8.6. 启动 YARN
##### 注意 #####：是在 mini03 上执行 start-yarn.sh，把 namenode 和 resourcemanager 分开是因为性能问题
# 因为他们都要占用大量资源，所以把他们分开了，他们分开了就要分别在不同的机器上启动
[yun@mini03 ~]$ start-yarn.sh
starting yarn daemons
starting resourcemanager, logging to /app/hadoop-2.7.6/logs/yarn-yun-resourcemanager-mini03.out
mini06: starting nodemanager, logging to /app/hadoop-2.7.6/logs/yarn-yun-nodemanager-mini06.out
mini07: starting nodemanager, logging to /app/hadoop-2.7.6/logs/yarn-yun-nodemanager-mini07.out
mini05: starting nodemanager, logging to /app/hadoop-2.7.6/logs/yarn-yun-nodemanager-mini05.out
################################
# 在 mini04 启动 resourcemanager
[yun@mini04 ~]$ yarn-daemon.sh start resourcemanager # 也可用 start-yarn.sh
starting resourcemanager, logging to /app/hadoop-2.7.6/logs/yarn-yun-resourcemanager-mini04.out
8.7. 启动说明
# 第一次启动的时候请严格按照上面的步骤【第一次涉及格式化问题】
# 第二次以及之后，步骤为：启动 zookeeper、HDFS、YARN
9. 浏览访问
9.1. Hdfs 访问
9.1.1. 正常情况访问
http://mini01:50070
http://mini02:50070
9.1.2. mini01 挂了 Active 自动切换
# mini01 操作
[yun@mini01 ~]$ jps
3584 DFSZKFailoverController
3283 NameNode
5831 Jps
[yun@mini01 ~]$ kill 3283
[yun@mini01 ~]$ jps
3584 DFSZKFailoverController
5893 Jps
Namenode 挂了所以 mini01 不能访问
http://mini02:50070
可见 Hadoop 已经切换过去了，之后 mini01 即使起来了，状态也只能为 standby。
9.2. Yarn 访问
http://mini03:8088
http://mini04:8088
会直接跳转到 http://mini03:8088/
# 该图从其他地方截取，所以不怎么匹配
# Linux 下访问
[yun@mini01 ~]$ curl mini04:8088
This is standby RM. The redirect url is: http://mini03:8088/
HA 完毕
10. 集群运维测试
10.1. Haadmin 与状态切换管理
[yun@mini01 ~]$ hdfs haadmin
Usage: haadmin
[-transitionToActive [–forceactive] <serviceId>]
[-transitionToStandby <serviceId>]
[-failover [–forcefence] [–forceactive] <serviceId> <serviceId>]
[-getServiceState <serviceId>]
[-checkHealth <serviceId>]
[-help <command>]
Generic options supported are
-conf <configuration file> specify an application configuration file
-D <property=value> use value for given property
-fs <local|namenode:port> specify a namenode
-jt <local|resourcemanager:port> specify a ResourceManager
-files <comma separated list of files> specify comma separated files to be copied to the map reduce cluster
-libjars <comma separated list of jars> specify comma separated jar files to include in the classpath.
-archives <comma separated list of archives> specify comma separated archives to be unarchived on the compute machines.
The general command line syntax is
bin/hadoop command [genericOptions] [commandOptions]
可以看到，状态操作的命令示例：
# 查看 namenode 工作状态
hdfs haadmin -getServiceState nn1
# 将 standby 状态 namenode 切换到 active
hdfs haadmin -transitionToActive nn1
# 将 active 状态 namenode 切换到 standby
hdfs haadmin -transitionToStandby nn2
10.2. 测试集群工作状态的一些指令
测试集群工作状态的一些指令：
hdfs dfsadmin -report 查看 hdfs 的各节点状态信息
hdfs haadmin -getServiceState nn1 # hdfs haadmin -getServiceState nn2 获取一个 namenode 节点的 HA 状态
hadoop-daemon.sh start namenode 单独启动一个 namenode 进程
hadoop-daemon.sh start zkfc 单独启动一个 zkfc 进程
10.3. Datanode 动态上下线
Datanode 动态上下线很简单，步骤如下：
a) 准备一台服务器，设置好环境
b) 部署 hadoop 的安装包，并同步集群配置
c) 联网上线，新 datanode 会自动加入集群
d) 如果是一次增加大批 datanode，还应该做集群负载重均衡
10.4. 数据块的 balance
启动 balancer 的命令：
start-balancer.sh -threshold 8
运行之后，会有 Balancer 进程出现：
上述命令设置了 Threshold 为 8%，那么执行 balancer 命令的时候，首先统计所有 DataNode 的磁盘利用率的均值，然后判断如果某一个 DataNode 的磁盘利用率超过这个均值 Threshold，那么将会把这个 DataNode 的 block 转移到磁盘利用率低的 DataNode，这对于新节点的加入来说十分有用。Threshold 的值为 1 到 100 之间，不显示的进行参数设置的话，默认是 10。
Hadoop2.3-HA 高可用集群环境搭建 https://www.linuxidc.com/Linux/2017-03/142155.htm
Hadoop 项目之基于 CentOS7 的 Cloudera 5.10.1（CDH）的安装部署 https://www.linuxidc.com/Linux/2017-04/143095.htm
Hadoop2.7.2 集群搭建详解（高可用）https://www.linuxidc.com/Linux/2017-03/142052.htm
使用 Ambari 来部署 Hadoop 集群（搭建内网 HDP 源）https://www.linuxidc.com/Linux/2017-03/142136.htm
Ubuntu 14.04 下 Hadoop 集群安装 https://www.linuxidc.com/Linux/2017-02/140783.htm
CentOS 6.7 安装 Hadoop 2.7.2 https://www.linuxidc.com/Linux/2017-08/146232.htm
Ubuntu 16.04 上构建分布式 Hadoop-2.7.3 集群 https://www.linuxidc.com/Linux/2017-07/145503.htm
CentOS 7 下 Hadoop 2.6.4 分布式集群环境搭建 https://www.linuxidc.com/Linux/2017-06/144932.htm
Hadoop2.7.3+Spark2.1.0 完全分布式集群搭建过程 https://www.linuxidc.com/Linux/2017-06/144926.htm
CentOS 7.4 下编译安装 Hadoop 2.7.6 及所需文件 https://www.linuxidc.com/Linux/2018-06/152786.htm
Ubuntu 16.04.3 下安装配置 Hadoop https://www.linuxidc.com/Linux/2018-04/151993.htm
更多 Hadoop 相关信息见 Hadoop 专题页面 http://www.linuxidc.com/topicnews.aspx?tid=13
：
