1.HDFS简单版集群搭建相关配置文件
1.core-site.xml文件
1
2 fs.defaultFS
3 hdfs://hadoop2:9000
4
5
6
7 hadoop.tmp.dir
8 /usr/hadoop-2.9.2/data
9
2.ZK搭建高可用HDFS集群搭建相关配置文件
1.zkdata1/zoo.cfg文件
1 tickTime=2000
2 initLimit=10
3 syncLimit=5
4 dataDir=/root/zkdata
5 clientPort=3001
6 server.1=主机名:3002:3003
7 server.2=主机名:4002:4003
8 server.3=主机名:5002:5003
2.hadoop的core-site.xml文件
1
2
3 fs.defaultFS
4 hdfs://ns
5
6
7
8 hadoop.tmp.dir
9 /usr/hadoop-2.9.2/data
10
11
12
13 ha.zookeeper.quorum
14 hadoop1:3001,hadoop1:4001,hadoop1:5001
15
3.hadoop 配置hdfs-site.xml文件
1
2
3 dfs.nameservices
4 ns
5
6
7
8 dfs.ha.namenodes.ns
9 nn1,nn2
10
11
12
13 dfs.namenode.rpc-address.ns.nn1
14 hadoop2:9000
15
16
17
18 dfs.namenode.http-address.ns.nn1
19 hadoop2:50070
20
21
22
23 dfs.namenode.rpc-address.ns.nn2
24 hadoop3:9000
25
26
27
28 dfs.namenode.http-address.ns.nn2
29 hadoop3:50070
30
31
32
33
34 dfs.namenode.shared.edits.dir
35 qjournal://hadoop2:8485;hadoop3:8485;hadoop4:8485/ns
36
37
38
39 dfs.journalnode.edits.dir
40 /root/journal
41
42
43
44 dfs.ha.automatic-failover.enabled
45 true
46
47
48
49 dfs.client.failover.proxy.provider.ns
50 org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider
51
52
53
54 dfs.ha.fencing.methods
55 sshfence
56
57
58
59 dfs.ha.fencing.ssh.private-key-files
60 /root/.ssh/id_rsa
61
3.搭建yarn集群
1.mapred-site.xml
注意:默认/etc/中没有这个配置文件 需要拷贝mapred-site.xml.template 配置文件
改名为mapred-site.xml
1
2 mapreduce.framework.name
3 yarn
4
2.yarn.site.xml文件
1
2 yarn.nodemanager.aux-services
3 mapreduce_shuffle
4
5
6 yarn.resourcemanager.hostname
7 Hadoop
8
4.HA的hadoop集群搭建的配置文件(最终版)
1.core-site.xml文件
1
2
3 fs.defaultFS
4 hdfs://ns
5
6
7
8 hadoop.tmp.dir
9 /root/hadoop-2.9.2/data
10
11
12
13 ha.zookeeper.quorum
14 zk:3001,zk:4001,zk:5001
15
2.hdfs-site.xml文件
1
2
3 dfs.nameservices
4 ns
5
6
7
8 dfs.ha.namenodes.ns
9 nn1,nn2
10
11
12
13 dfs.namenode.rpc-address.ns.nn1
14 hadoop22:9000
15
16
17
18 dfs.namenode.http-address.ns.nn1
19 hadoop22:50070
20
21
22
23 dfs.namenode.rpc-address.ns.nn2
24 hadoop23:9000
25
26
27
28 dfs.namenode.http-address.ns.nn2
29 hadoop23:50070
30
31
32
33
34 dfs.namenode.shared.edits.dir
35 qjournal://hadoop26:8485;hadoop27:8485;hadoop28:8485/ns
36
37
38
39 dfs.journalnode.edits.dir
40 /root/journal
41
42
43
44 dfs.ha.automatic-failover.enabled
45 true
46
47
48
49 dfs.client.failover.proxy.provider.ns
50 org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider
51
52
53
54 dfs.ha.fencing.methods
55 sshfence
56
57
58
59 dfs.ha.fencing.ssh.private-key-files
60 /root/.ssh/id_rsa
61
3.yarn-site.xml文件
1
2
3 yarn.resourcemanager.ha.enabled
4 true
5
6
7
8 yarn.resourcemanager.cluster-id
9 yrc
10
11
12
13 yarn.resourcemanager.ha.rm-ids
14 rm1,rm2
15
16
17
18 yarn.resourcemanager.hostname.rm1
19 hadoop24
20
21
22 yarn.resourcemanager.hostname.rm2
23 hadoop25
24
25
26 yarn.resourcemanager.webapp.address.rm1
27 hadoop24:8088
28
29
30 yarn.resourcemanager.webapp.address.rm2
31 hadoop25:8088
32
33
34
35 yarn.resourcemanager.zk-address
36 zk:3001,zk:4001,zk:5001
37
38
39 yarn.nodemanager.aux-services
40 mapreduce_shuffle
41
4.mapred-site.xml 默认不存在需要复制
1
2
3 mapreduce.framework.name
4 yarn
5