文章目录
- 1、HA 概述
- 2、HDFS-HA 集群搭建
- 3、HDFS-HA 核心问题
- 4、HDFS-HA 手动模式
- 4.1 环境准备
- 4.2 规划集群
- 4.3 配置 HDFS-HA 集群
- 4.4 启动 HDFS-HA 集群
- 5、HDFS-HA 自动模式
- 5.1 HDFS-HA 自动故障转移工作机制
- 5.2 HDFS-HA 自动故障转移的集群规划
- 5.3 配置 HDFS-HA 自动故障转移
- 5.4 上传文件演示
- 5.5 解决 NN 连接不上 JN 的问题
- 6、Yarn-HA配置
- 6.1 YARN-HA 工作机制
- 6.2 配置 YARN-HA 集群
- 6.3 HADOOP HA 的最终规划
- 7、HDFS Federation架构设计
1、HA 概述
(1)所谓 HA(High Availablity),即高可用(7*24 小时不中断服务)。
(2)实现高可用最关键的策略是消除单点故障。HA 严格来说应该分成各个组件的 HA
机制:HDFS 的 HA 和 YARN 的 HA。
(3)NameNode 主要在以下两个方面影响 HDFS 集群
➢ NameNode 机器发生意外,如宕机,集群将无法使用,直到管理员重启
➢ NameNode 机器需要升级,包括软件、硬件升级,此时集群也将无法使用
HDFS HA 功能通过配置多个 NameNodes(Active/Standby)实现在集群中对 NameNode 的
热备来解决上述问题。如果出现故障,如机器崩溃或机器需要升级维护,这时可通过此种方
式将 NameNode 很快的切换到另外一台机器。
2、HDFS-HA 集群搭建
3、HDFS-HA 核心问题
1)怎么保证三台 namenode 的数据一致
a.Fsimage:让一台 nn 生成数据,让其他机器 nn 同步
b.Edits:需要引进新的模块 JournalNode 来保证 edtis 的文件的数据一致性
2)怎么让同时只有一台 nn 是 active,其他所有是 standby 的
a.手动分配
b.自动分配
3)2nn 在 ha 架构中并不存在,定期合并 fsimage 和 edtis 的活谁来干
由 standby 的 nn 来干
4)如果 nn 真的发生了问题,怎么让其他的 nn 上位干活
a.手动故障转移
b.自动故障转移
4、HDFS-HA 手动模式
4.1 环境准备
(1)修改 IP
(2)修改主机名及主机名和 IP 地址的映射
(3)关闭防火墙
(4)ssh 免密登录
(5)安装 JDK,配置环境变量等
4.2 规划集群
4.3 配置 HDFS-HA 集群
1)官方地址:http://hadoop.apache.org/
2)在 opt 目录下创建一个 ha 文件夹
[atguigu@hadoop102 ~]$ cd /opt
[atguigu@hadoop102 opt]$ sudo mkdir ha
[atguigu@hadoop102 opt]$ sudo chown atguigu:atguigu /opt/ha
3)将/opt/module/下的 hadoop-3.1.3 拷贝到/opt/ha 目录下(记得删除 data 和 log 目录)
[atguigu@hadoop102 opt]$ cp -r /opt/module/hadoop-3.1.3 /opt/ha/
4)配置 core-site.xml
<configuration>
<!-- 把多个 NameNode 的地址组装成一个集群 mycluster -->
<property>
<name>fs.defaultFS</name>
<value>hdfs://mycluster</value>
</property>
<!-- 指定 hadoop 运行时产生文件的存储目录 -->
<property>
<name>hadoop.tmp.dir</name>
<value>/opt/ha/hadoop-3.1.3/data</value>
</property>
</configuration>
5)配置 hdfs-site.xml
<configuration>
<!-- NameNode 数据存储目录 -->
<property>
<name>dfs.namenode.name.dir</name>
<value>file://${hadoop.tmp.dir}/name</value>
</property>
<!-- DataNode 数据存储目录 -->
<property>
<name>dfs.datanode.data.dir</name>
<value>file://${hadoop.tmp.dir}/data</value>
</property>
<!-- JournalNode 数据存储目录 -->
<property>
<name>dfs.journalnode.edits.dir</name>
<value>${hadoop.tmp.dir}/jn</value>
</property>
<!-- 完全分布式集群名称 -->
<property>
<name>dfs.nameservices</name>
<value>mycluster</value>
</property>
<!-- 集群中 NameNode 节点都有哪些 -->
<property>
<name>dfs.ha.namenodes.mycluster</name>
<value>nn1,nn2,nn3</value>
</property>
<!-- NameNode 的 RPC 通信地址 -->
<property>
<name>dfs.namenode.rpc-address.mycluster.nn1</name>
<value>hadoop102:8020</value>
</property>
<property>
<name>dfs.namenode.rpc-address.mycluster.nn2</name>
<value>hadoop103:8020</value>
</property>
<property>
<name>dfs.namenode.rpc-address.mycluster.nn3</name>
<value>hadoop104:8020</value>
</property>
<!-- NameNode 的 http 通信地址 -->
<property>
<name>dfs.namenode.http-address.mycluster.nn1</name>
<value>hadoop102:9870</value>
</property>
<property>
<name>dfs.namenode.http-address.mycluster.nn2</name>
<value>hadoop103:9870</value>
</property>
<property>
<name>dfs.namenode.http-address.mycluster.nn3</name>
<value>hadoop104:9870</value>
</property>
<!-- 指定 NameNode 元数据在 JournalNode 上的存放位置 -->
<property>
<name>dfs.namenode.shared.edits.dir</name>
<value>qjournal://hadoop102:8485;hadoop103:8485;hadoop104:8485/mycluster</value>
</property>
<!-- 访问代理类:client 用于确定哪个 NameNode 为 Active -->
<property>
<name>dfs.client.failover.proxy.provider.mycluster</name>
<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
</property>
<!-- 配置隔离机制,即同一时刻只能有一台服务器对外响应 -->
<property>
<name>dfs.ha.fencing.methods</name>
<value>sshfence</value>
</property>
<!-- 使用隔离机制时需要 ssh 秘钥登录-->
<property>
<name>dfs.ha.fencing.ssh.private-key-files</name>
<value>/home/atguigu/.ssh/id_rsa</value>
</property>
</configuration>
6)分发配置好的 hadoop 环境到其他节点
4.4 启动 HDFS-HA 集群
1)将 HADOOP_HOME 环境变量更改到 HA 目录(三台机器)
[atguigu@hadoop102 ~]$ sudo vim /etc/profile.d/my_env.sh
将 HADOOP_HOME 部分改为如下
#HADOOP_HOME
export HADOOP_HOME=/opt/ha/hadoop-3.1.3
export PATH=$PATH:$HADOOP_HOME/bin
export PATH=$PATH:$HADOOP_HOME/sbin
去三台机器上 source 环境变量
[atguigu@hadoop102 ~]$source /etc/profile
2)在各个 JournalNode 节点上,输入以下命令启动 journalnode 服务
[atguigu@hadoop102 ~]$ hdfs --daemon start journalnode
[atguigu@hadoop103 ~]$ hdfs --daemon start journalnode
[atguigu@hadoop104 ~]$ hdfs --daemon start journalnode
3)在[nn1]上,对其进行格式化,并启动
[atguigu@hadoop102 ~]$ hdfs namenode -format
[atguigu@hadoop102 ~]$ hdfs --daemon start namenode
4)在[nn2]和[nn3]上,同步 nn1 的元数据信息
[atguigu@hadoop103 ~]$ hdfs namenode -bootstrapStandby
[atguigu@hadoop104 ~]$ hdfs namenode -bootstrapStandby
5)启动[nn2]和[nn3]
[atguigu@hadoop103 ~]$ hdfs --daemon start namenode
[atguigu@hadoop104 ~]$ hdfs --daemon start namenode
6)查看 web 页面显示
7)在所有节点上,启动 datanode
[atguigu@hadoop102 ~]$ hdfs --daemon start datanode
[atguigu@hadoop103 ~]$ hdfs --daemon start datanode
[atguigu@hadoop104 ~]$ hdfs --daemon start datanode
8)将[nn1]切换为 Active
[atguigu@hadoop102 ~]$ hdfs haadmin -transitionToActive nn1
9)查看是否 Active
[atguigu@hadoop102 ~]$ hdfs haadmin -getServiceState nn1
5、HDFS-HA 自动模式
5.1 HDFS-HA 自动故障转移工作机制
自动故障转移为 HDFS 部署增加了两个新组件:ZooKeeper 和 ZKFailoverController
(ZKFC)进程,如图所示。ZooKeeper 是维护少量协调数据,通知客户端这些数据的改变
和监视客户端故障的高可用服务。
5.2 HDFS-HA 自动故障转移的集群规划
5.3 配置 HDFS-HA 自动故障转移
1)具体配置
(1)在 hdfs-site.xml 中增加
<!-- 启用 nn 故障自动转移 -->
<property>
<name>dfs.ha.automatic-failover.enabled</name>
<value>true</value>
</property>
(2)在 core-site.xml 文件中增加
<!-- 指定 zkfc 要连接的 zkServer 地址 -->
<property>
<name>ha.zookeeper.quorum</name>
<value>hadoop102:2181,hadoop103:2181,hadoop104:2181</value>
</property>
(3)修改后分发配置文件
[atguigu@hadoop102 etc]$ pwd
/opt/ha/hadoop-3.1.3/etc
[atguigu@hadoop102 etc]$ xsync hadoop/
2)启动
(1)关闭所有 HDFS 服务:
[atguigu@hadoop102 ~]$ stop-dfs.sh
(2)启动 Zookeeper 集群:
[atguigu@hadoop102 ~]$ zkServer.sh start
[atguigu@hadoop103 ~]$ zkServer.sh start
[atguigu@hadoop104 ~]$ zkServer.sh start
(3)启动 Zookeeper 以后,然后再初始化 HA 在 Zookeeper 中状态:
[atguigu@hadoop102 ~]$ hdfs zkfc -formatZK
(4)启动 HDFS 服务:
[atguigu@hadoop102 ~]$ start-dfs.sh
(5)可以去 zkCli.sh 客户端查看 Namenode 选举锁节点内容:
[zk: localhost:2181(CONNECTED) 7] get -s
/hadoop-ha/mycluster/ActiveStandbyElectorLock
myclusternn2 hadoop103 �>(�>
cZxid = 0x10000000b
ctime = Tue Jul 14 17:00:13 CST 2020
mZxid = 0x10000000b
mtime = Tue Jul 14 17:00:13 CST 2020
pZxid = 0x10000000b
cversion = 0
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x40000da2eb70000
dataLength = 33
numChildren = 0
3)验证
将 Active NameNode 进程 kill,查看网页端三台 Namenode 的状态变化
[atguigu@hadoop102 ~]$ kill -9 namenode 的进程 id
5.4 上传文件演示
hadoop fs -put test.txt /
hadoop fs -put test.txt http://mycluster/
5.5 解决 NN 连接不上 JN 的问题
自动故障转移配置好以后,然后使用 start-dfs.sh 群起脚本启动 hdfs 集群,有可能会遇到 NameNode 起来一会后,进程自动关闭的问题。查看 NameNode 日志,报错信息如下:
2020-08-17 10:11:40,658 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop104/192.168.6.104:8485. Already tried 0 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:40,659 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop102/192.168.6.102:8485. Already tried 0 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:40,659 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop103/192.168.6.103:8485. Already tried 0 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:41,660 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop104/192.168.6.104:8485. Already tried 1 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:41,660 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop102/192.168.6.102:8485. Already tried 1 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:41,665 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop103/192.168.6.103:8485. Already tried 1 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:42,661 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop104/192.168.6.104:8485. Already tried 2 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:42,661 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop102/192.168.6.102:8485. Already tried 2 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:42,667 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop103/192.168.6.103:8485. Already tried 2 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:43,662 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop104/192.168.6.104:8485. Already tried 3 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:43,662 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop102/192.168.6.102:8485. Already tried 3 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:43,668 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop103/192.168.6.103:8485. Already tried 3 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:44,663 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop104/192.168.6.104:8485. Already tried 4 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:44,663 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop102/192.168.6.102:8485. Already tried 4 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:44,670 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop103/192.168.6.103:8485. Already tried 4 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:45,467 INFO
org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 6001
ms (timeout=20000 ms) for a response for selectStreamingInputStreams. No
responses yet.
2020-08-17 10:11:45,664 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop102/192.168.6.102:8485. Already tried 5 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:45,664 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop104/192.168.6.104:8485. Already tried 5 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:45,672 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop103/192.168.6.103:8485. Already tried 5 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:46,469 INFO
org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 7003
ms (timeout=20000 ms) for a response for selectStreamingInputStreams. No
responses yet.
2020-08-17 10:11:46,665 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop102/192.168.6.102:8485. Already tried 6 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:46,665 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop104/192.168.6.104:8485. Already tried 6 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:46,673 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop103/192.168.6.103:8485. Already tried 6 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:47,470 INFO
org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 8004
ms (timeout=20000 ms) for a response for selectStreamingInputStreams. No
responses yet.
2020-08-17 10:11:47,666 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop102/192.168.6.102:8485. Already tried 7 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:47,667 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop104/192.168.6.104:8485. Already tried 7 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:47,674 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop103/192.168.6.103:8485. Already tried 7 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:48,471 INFO
org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager: Waited 9005
ms (timeout=20000 ms) for a response for selectStreamingInputStreams. No
responses yet.
2020-08-17 10:11:48,668 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop102/192.168.6.102:8485. Already tried 8 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:48,668 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop104/192.168.6.104:8485. Already tried 8 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:48,675 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop103/192.168.6.103:8485. Already tried 8 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:49,669 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop102/192.168.6.102:8485. Already tried 9 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:49,673 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop104/192.168.6.104:8485. Already tried 9 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:49,676 INFO org.apache.hadoop.ipc.Client: Retrying connect
to server: hadoop103/192.168.6.103:8485. Already tried 9 time(s); retry
policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10,
sleepTime=1000 MILLISECONDS)
2020-08-17 10:11:49,678 WARN
org.apache.hadoop.hdfs.server.namenode.FSEditLog: Unable to determine input
streams from QJM to [192.168.6.102:8485, 192.168.6.103:8485,
192.168.6.104:8485]. Skipping.
org.apache.hadoop.hdfs.qjournal.client.QuorumException: Got too many
exceptions to achieve quorum size 2/3. 3 exceptions thrown:
192.168.6.103:8485: Call From hadoop102/192.168.6.102 to hadoop103:8485
failed on connection exception: java.net.ConnectException: 拒绝连接; For more
details see: http://wiki.apache.org/hadoop/ConnectionRefused
192.168.6.102:8485: Call From hadoop102/192.168.6.102 to hadoop102:8485
failed on connection exception: java.net.ConnectException: 拒绝连接; For more
details see: http://wiki.apache.org/hadoop/ConnectionRefused
192.168.6.104:8485: Call From hadoop102/192.168.6.102 to hadoop104:8485
failed on connection exception: java.net.ConnectException: 拒绝连接; For more
details see: http://wiki.apache.org/hadoop/ConnectionRefused
查看报错日志,可分析出报错原因是因为 NameNode 连接不上 JournalNode,而利
用 jps 命令查看到三台 JN 都已经正常启动,为什么 NN 还是无法正常连接到 JN 呢?这
是因为 start-dfs.sh 群起脚本默认的启动顺序是先启动 NN,再启动 DN,然后再启动 JN,
并且默认的 rpc 连接参数是重试次数为 10,每次重试的间隔是 1s,也就是说启动完 NN
以后的 10s 中内,JN 还启动不起来,NN 就会报错了。
core-default.xml 里面有两个参数如下:
<!-- NN 连接 JN 重试次数,默认是 10 次 -->
<property>
<name>ipc.client.connect.max.retries</name>
<value>10</value>
</property>
<!-- 重试时间间隔,默认 1s -->
<property>
<name>ipc.client.connect.retry.interval</name>
<value>1000</value>
</property>
解决方案:遇到上述问题后,可以稍等片刻,等 JN 成功启动后,手动启动下三台
NN:
[atguigu@hadoop102 ~]$ hdfs --daemon start namenode
[atguigu@hadoop103 ~]$ hdfs --daemon start namenode
[atguigu@hadoop104 ~]$ hdfs --daemon start namenode
也可以在 core-site.xml 里面适当调大上面的两个参数:
<!-- NN 连接 JN 重试次数,默认是 10 次 -->
<property>
<name>ipc.client.connect.max.retries</name>
<value>20</value>
</property>
<!-- 重试时间间隔,默认 1s -->
<property>
<name>ipc.client.connect.retry.interval</name>
<value>5000</value>
</property>
6、Yarn-HA配置
6.1 YARN-HA 工作机制
1)官方文档:
http://hadoop.apache.org/docs/r3.1.3/hadoop-yarn/hadoop-yarn-site/ResourceManagerHA.html
2)YARN-HA 工作机制
6.2 配置 YARN-HA 集群
1)环境准备
(1)修改 IP
(2)修改主机名及主机名和 IP 地址的映射
(3)关闭防火墙
(4)ssh 免密登录
(5)安装 JDK,配置环境变量等
(6)配置 Zookeeper 集群
2)规划集群
3)核心问题
a .如果当前 active rm 挂了,其他 rm 怎么将其他 standby rm 上位
核心原理跟 hdfs 一样,利用了 zk 的临时节点
b. 当前 rm 上有很多的计算程序在等待运行,其他的 rm 怎么将这些程序接手过来接着跑
rm 会将当前的所有计算程序的状态存储在 zk 中,其他 rm 上位后会去读取,然后接着跑
4)具体配置
(1)yarn-site.xml
<configuration>
<property>
<name>yarn.nodemanager.aux-services</name>
<value>mapreduce_shuffle</value>
</property>
<!-- 启用 resourcemanager ha -->
<property>
<name>yarn.resourcemanager.ha.enabled</name>
<value>true</value>
</property>
<!-- 声明 resourcemanager 的地址 -->
<property>
<name>yarn.resourcemanager.cluster-id</name>
<value>cluster-yarn1</value>
</property>
<!--指定 resourcemanager 的逻辑列表-->
<property>
<name>yarn.resourcemanager.ha.rm-ids</name>
<value>rm1,rm2,rm3</value>
</property>
<!-- ========== rm1 的配置 ========== -->
<!-- 指定 rm1 的主机名 -->
<property>
<name>yarn.resourcemanager.hostname.rm1</name>
<value>hadoop102</value>
</property>
<!-- 指定 rm1 的 web 端地址 -->
<property>
<name>yarn.resourcemanager.webapp.address.rm1</name>
<value>hadoop102:8088</value>
</property>
<!-- 指定 rm1 的内部通信地址 -->
<property>
<name>yarn.resourcemanager.address.rm1</name>
<value>hadoop102:8032</value>
</property>
<!-- 指定 AM 向 rm1 申请资源的地址 -->
<property>
<name>yarn.resourcemanager.scheduler.address.rm1</name>
<value>hadoop102:8030</value>
</property>
<!-- 指定供 NM 连接的地址 -->
<property>
<name>yarn.resourcemanager.resource-tracker.address.rm1</name>
<value>hadoop102:8031</value>
</property>
<!-- ========== rm2 的配置 ========== -->
<!-- 指定 rm2 的主机名 -->
<property>
<name>yarn.resourcemanager.hostname.rm2</name>
<value>hadoop103</value>
</property>
<property>
<name>yarn.resourcemanager.webapp.address.rm2</name>
<value>hadoop103:8088</value>
</property>
<property>
<name>yarn.resourcemanager.address.rm2</name>
<value>hadoop103:8032</value>
</property>
<property>
<name>yarn.resourcemanager.scheduler.address.rm2</name>
<value>hadoop103:8030</value>
</property>
<property>
<name>yarn.resourcemanager.resource-tracker.address.rm2</name>
<value>hadoop103:8031</value>
</property>
<!-- ========== rm3 的配置 ========== -->
<!-- 指定 rm1 的主机名 -->
<property>
<name>yarn.resourcemanager.hostname.rm3</name>
<value>hadoop104</value>
</property>
<!-- 指定 rm1 的 web 端地址 -->
<property>
<name>yarn.resourcemanager.webapp.address.rm3</name>
<value>hadoop104:8088</value>
</property>
<!-- 指定 rm1 的内部通信地址 -->
<property>
<name>yarn.resourcemanager.address.rm3</name>
<value>hadoop104:8032</value>
</property>
<!-- 指定 AM 向 rm1 申请资源的地址 -->
<property>
<name>yarn.resourcemanager.scheduler.address.rm3</name>
<value>hadoop104:8030</value>
</property>
<!-- 指定供 NM 连接的地址 -->
<property>
<name>yarn.resourcemanager.resource-tracker.address.rm3</name>
<value>hadoop104:8031</value>
</property>
<!-- 指定 zookeeper 集群的地址 -->
<property>
<name>yarn.resourcemanager.zk-address</name>
<value>hadoop102:2181,hadoop103:2181,hadoop104:2181</value>
</property>
<!-- 启用自动恢复 -->
<property>
<name>yarn.resourcemanager.recovery.enabled</name>
<value>true</value>
</property>
<!-- 指定 resourcemanager 的状态信息存储在 zookeeper 集群 -->
<property>
<name>yarn.resourcemanager.store.class</name>
<value>org.apache.hadoop.yarn.server.resourcemanager.recovery.ZKRMStateStore</value>
</property>
<!-- 环境变量的继承 -->
<property>
<name>yarn.nodemanager.env-whitelist</name>
<value>JAVA_HOME,HADOOP_COMMON_HOME,HADOOP_HDFS_HOME,HADOOP_CONF_DIR,CLASSPATH_PREPEND_DISTCACHE,HADOOP_YARN_HOME,HADOOP_MAPRED_HOME</value>
</property>
</configuration>
(2)同步更新其他节点的配置信息,分发配置文件
[atguigu@hadoop102 etc]$ xsync hadoop/
4)启动 YARN
(1)在 hadoop102 或者 hadoop103 中执行:
[atguigu@hadoop102 ~]$ start-yarn.sh
(2)查看服务状态
[atguigu@hadoop102 ~]$ yarn rmadmin -getServiceState rm1
(3)可以去 zkCli.sh 客户端查看 ResourceManager 选举锁节点内容:
[atguigu@hadoop102 ~]$ zkCli.sh
[zk: localhost:2181(CONNECTED) 16] get -s
/yarn-leader-election/cluster-yarn1/ActiveStandbyElectorLock
cluster-yarn1rm1
cZxid = 0x100000022
ctime = Tue Jul 14 17:06:44 CST 2020
mZxid = 0x100000022
mtime = Tue Jul 14 17:06:44 CST 2020
pZxid = 0x100000022
cversion = 0
dataVersion = 0
aclVersion = 0
ephemeralOwner = 0x30000da33080005
dataLength = 20
numChildren = 0
(4)web 端查看 hadoop102:8088 和 hadoop103:8088 的 YARN 的状态
6.3 HADOOP HA 的最终规划
将整个 ha 搭建完成后,集群将形成以下模样
7、HDFS Federation架构设计
- NameNode架构的局限性
(1)Namespace(命名空间)的限制
由于NameNode在内存中存储所有的元数据(metadata),因此单个NameNode所能存储的对象(文件+块)数目受到NameNode所在JVM的heap size的限制。50G的heap能够存储20亿(200million)个对象,这20亿个对象支持4000个DataNode,12PB的存储(假设文件平均大小为40MB)。随着数据的飞速增长,存储的需求也随之增长。单个DataNode从4T增长到36T,集群的尺寸增长到8000个DataNode。存储的需求从12PB增长到大于100PB。
(2)隔离问题
由于HDFS仅有一个NameNode,无法隔离各个程序,因此HDFS上的一个实验程序就很有可能影响整个HDFS上运行的程序。
(3)性能的瓶颈
由于是单个NameNode的HDFS架构,因此整个HDFS文件系统的吞吐量受限于单个NameNode的吞吐量。
- HDFS Federation架构设计,如图所示
能不能有多个NameNode
- HDFS Federation应用思考
不同应用可以使用不同NameNode进行数据管理
图片业务、爬虫业务、日志审计业务
Hadoop生态系统中,不同的框架使用不同的NameNode进行管理NameSpace。(隔离性)