卸载Ambari集群

清理ambari安装的hadoop集群

本文针对redhat或者centos

对于测试集群，如果通过ambari安装hadoop集群后，想重新再来一次的话，需要清理集群。

对于安装了很多hadoop组件的话，这个工作很繁琐。接下来是我整理的清理过程。

1，通过ambari将集群中的所用组件都关闭，如果关闭不了，直接kill -9 XXX

2，关闭ambari-server，ambari-agent

1. ambari-server stop
2. ambari-agent stop

3，卸载安装的软件

1. yum remove hadoop_2* hdp-select* ranger_2* zookeeper* bigtop* atlas-metadata* ambari* postgresql spark*

以上命令可能不全，执行完一下命令后，再执行

1. yum list | grep @HDP

查看是否还有没有卸载的，如果有，继续通过#yum remove XXX卸载

4，删除postgresql的数据

postgresql软件卸载后，其数据还保留在硬盘中，需要把这部分数据删除掉，如果不删除掉，重新安装ambari-server后，有可能还应用以前的安装数据，而这些数据时错误数据，所以需要删除掉。

1. rm -rf /var/lib/pgsql

5，删除用户

ambari安装hadoop集群会创建一些用户，清除集群时有必要清除这些用户，并删除对应的文件夹。这样做可以避免集群运行时出现的文件访问权限错误的问题。

1. userdel oozie
2. userdel hive
3. userdel ambari-qa
4. userdel flume
5. userdel hdfs
6. userdel knox
7. userdel storm
8. userdel mapred
9. userdel hbase
10. userdel tez
11. userdel zookeeper
12. userdel kafka
13. userdel falcon
14. userdel sqoop
15. userdel yarn
16. userdel hcat
17. userdel atlas
18. userdel spark
19. rm -rf /home/atlas
20. rm -rf /home/accumulo
21. rm -rf /home/hbase
22. rm -rf /home/hive
23. rm -rf /home/oozie
24. rm -rf /home/storm
25. rm -rf /home/yarn
26. rm -rf /home/ambari-qa
27. rm -rf /home/falcon
28. rm -rf /home/hcat
29. rm -rf /home/kafka
30. rm -rf /home/mahout
31. rm -rf /home/spark
32. rm -rf /home/tez
33. rm -rf /home/zookeeper
34. rm -rf /home/flume
35. rm -rf /home/hdfs
36. rm -rf /home/knox
37. rm -rf /home/mapred
38. rm -rf /home/sqoop

6，删除ambari遗留数据

1. rm -rf /var/lib/ambari*
2. rm -rf /usr/lib/python2.6/site-packages/ambari_*
3. rm -rf /usr/lib/python2.6/site-packages/resource_management
4. rm -rf /usr/lib/ambari-*
5. rm -rf /etc/ambari-*

7，删除其他hadoop组件遗留数据

1. m -rf /etc/hadoop
2. rm -rf /etc/hbase
3. rm -rf /etc/hive
4. rm -rf /etc/oozie
5. rm -rf /etc/sqoop
6. rm -rf /etc/zookeeper
7. rm -rf /etc/flume
8. rm -rf /etc/storm
9. rm -rf /etc/hive-hcatalog
10. rm -rf /etc/tez
11. rm -rf /etc/falcon
12. rm -rf /etc/knox
13. rm -rf /etc/hive-webhcat
14. rm -rf /etc/kafka
15. rm -rf /etc/slider
16. rm -rf /etc/storm-slider-client
17. rm -rf /etc/spark
18. rm -rf /var/run/spark
19. rm -rf /var/run/hadoop
20. rm -rf /var/run/hbase
21. rm -rf /var/run/zookeeper
22. rm -rf /var/run/flume
23. rm -rf /var/run/storm
24. rm -rf /var/run/webhcat
25. rm -rf /var/run/hadoop-yarn
26. rm -rf /var/run/hadoop-mapreduce
27. rm -rf /var/run/kafka
28. rm -rf /var/log/hadoop
29. rm -rf /var/log/hbase
30. rm -rf /var/log/flume
31. rm -rf /var/log/storm
32. rm -rf /var/log/hadoop-yarn
33. rm -rf /var/log/hadoop-mapreduce
34. rm -rf /var/log/knox
35. rm -rf /usr/lib/flume
36. rm -rf /usr/lib/storm
37. rm -rf /var/lib/hive
38. rm -rf /var/lib/oozie
39. rm -rf /var/lib/flume
40. rm -rf /var/lib/hadoop-hdfs
41. rm -rf /var/lib/knox
42. rm -rf /var/log/hive
43. rm -rf /var/log/oozie
44. rm -rf /var/log/zookeeper
45. rm -rf /var/log/falcon
46. rm -rf /var/log/webhcat
47. rm -rf /var/log/spark
48. rm -rf /var/tmp/oozie
49. rm -rf /tmp/ambari-qa
50. rm -rf /var/hadoop
51. rm -rf /hadoop/falcon
52. rm -rf /tmp/hadoop
53. rm -rf /tmp/hadoop-hdfs
54. rm -rf /usr/hdp
55. rm -rf /usr/hadoop
56. rm -rf /opt/hadoop
57. rm -rf /tmp/hadoop
58. rm -rf /var/hadoop
59. rm -rf /hadoop
60. userdel nagios
61. userdel hive
62. userdel ambari-qa
63. userdel hbase
64. userdel oozie
65. userdel hcat
66. userdel mapred
67. userdel hdfs
68. userdel rrdcached
69. userdel zookeeper
70. userdel sqoop
71. userdel puppet
72. 可以查看：http://www.cnblogs.com/cenyuhai/p/3287855.html
73. userdel nagios
74. userdel hive
75. userdel ambari-qa
76. userdel hbase
77. userdel oozie
78. userdel hcat
79. userdel mapred
80. userdel hdfs
81. userdel rrdcached
82. userdel zookeeper
83. userdel sqoop
84. userdel puppet
85. userdel nagios
86. userdel hive
87. userdel ambari-qa
88. userdel hbase
89. userdel oozie
90. userdel hcat
91. userdel mapred
92. userdel hdfs
93. userdel rrdcached
94. userdel zookeeper
95. userdel sqoop
96. userdel puppet
97. userdel -r accumulo
98. userdel -r ambari-qa
99. userdel -r ams
100. userdel -r falcon
101. userdel -r flume
102. userdel -r hbase
103. userdel -r hcat
104. userdel -r hdfs
105. userdel -r hive
106. userdel -r kafka
107. userdel -r knox
108. userdel -r mapred
109. userdel -r oozie
110. userdel -r ranger
111. userdel -r spark
112. userdel -r sqoop
113. userdel -r storm
114. userdel -r tez
115. userdel -r yarn
116. userdel -r zeppelin
117. userdel -r zookeeper
118. userdel -r kms
119. userdel -r sqoop2
120. userdel -r hue
121. rm -rf /var/lib/hbase
122. rm -rf /var/lib/hive
123. rm -rf /var/lib/oozie
124. rm -rf /var/lib/sqoop
125. rm -rf /var/lib/zookeeper
126. rm -rf /var/lib/hadoop-hdfs
127. rm -rf /var/lib/hadoop-mapreduce

8，清理yum数据源

1. #yum clean all

通过以上清理后，重新安装ambari和hadoop集群（包括HDFS，YARN+MapReduce2，Zookeeper，Ambari Metrics，Spark）成功。如果安装其他组件碰到由于未清理彻底而导致的问题，请留言指出需要清理的数据，本人会补全该文档。

然后后续的重新搭建的方式，可以参考ambari2.6.0搭建文档。