Ambari学习笔记-删除ambari

1 在所有集群节点上运行python脚本

python /usr/lib/ambari-agent/lib/ambari_agent/HostCleanup.py --silent --skip=users

2 删除所有节点上的Hadoop包 

yum remove -y hive* 
yum remove -y oozie* 
yum remove -y pig* 
yum remove -y zookeeper* 
yum remove -y tez* 
yum remove -y hbase* 
yum remove -y ranger* 
yum remove -y knox* 
yum remove -y storm* 
yum remove -y accumulo* 
yum remove -y falcon* 
yum remove -y ambari-metrics-hadoop-sink 
yum remove -y smartsense-hst 
yum remove -y slider_2_4_2_0_258 
yum remove -y ambari-metrics-monitor 
yum remove -y spark2_2_5_3_0_37-yarn-shuffle 
yum remove -y spark_2_5_3_0_37-yarn-shuffle 
yum remove -y ambari-infra-solr-client 

3 删除ambari-agent

ambari-agent stop 
yum remove -y ambari-agent

4 删除ambari-server

ambari-server stop 
yum remove -y ambari-server 

5 删除所有节点上的存储库 

rm -rf /etc/yum.repos.d/ambari.repo /etc/yum.repos.d/HDP* 
yum clean all 

6 删除所有节点上的日志文件夹 

rm -rf /var/log/ambari-agent 
rm -rf /var/log/ambari-metrics-grafana 
rm -rf /var/log/ambari-metrics-monitor 
rm -rf /var/log/ambari-server/ 
rm -rf /var/log/falcon 
rm -rf /var/log/flume 
rm -rf /var/log/hadoop 
rm -rf /var/log/hadoop-mapreduce 
rm -rf /var/log/hadoop-yarn 
rm -rf /var/log/hive 
rm -rf /var/log/hive-hcatalog 
rm -rf /var/log/hive2 
rm -rf /var/log/hst 
rm -rf /var/log/knox 
rm -rf /var/log/oozie 
rm -rf /var/log/solr 
rm -rf /var/log/zookeeper 

7 删除所有节点上的Hadoop文件夹，包括HDFS数据 

rm -rf /hadoop/* 
rm -rf /hdfs/hadoop 
rm -rf /hdfs/lost+found 
rm -rf /hdfs/var 
rm -rf /local/opt/hadoop 
rm -rf /tmp/hadoop 
rm -rf /usr/bin/hadoop 
rm -rf /usr/hdp 
rm -rf /var/hadoop 

8 删除所有节点上的配置文件夹 

rm -rf /etc/ambari-agent 
rm -rf /etc/ambari-metrics-grafana 
rm -rf /etc/ambari-server 
rm -rf /etc/ams-hbase 
rm -rf /etc/falcon 
rm -rf /etc/flume 
rm -rf /etc/hadoop 
rm -rf /etc/hadoop-httpfs 
rm -rf /etc/hbase 
rm -rf /etc/hive 
rm -rf /etc/hive-hcatalog 
rm -rf /etc/hive-webhcat 
rm -rf /etc/hive2 
rm -rf /etc/hst 
rm -rf /etc/knox 
rm -rf /etc/livy 
rm -rf /etc/mahout 
rm -rf /etc/oozie 
rm -rf /etc/phoenix 
rm -rf /etc/pig 
rm -rf /etc/ranger-admin 
rm -rf /etc/ranger-usersync 
rm -rf /etc/spark2 
rm -rf /etc/tez 
rm -rf /etc/tez_hive2 
rm -rf /etc/zookeeper

10 删除所有节点上的PID

rm -rf /var/run/ambari-agent 
rm -rf /var/run/ambari-metrics-grafana 
rm -rf /var/run/ambari-server 
rm -rf /var/run/falcon 
rm -rf /var/run/flume 
rm -rf /var/run/hadoop 
rm -rf /var/run/hadoop-mapreduce 
rm -rf /var/run/hadoop-yarn 
rm -rf /var/run/hbase 
rm -rf /var/run/hive 
rm -rf /var/run/hive-hcatalog 
rm -rf /var/run/hive2 
rm -rf /var/run/hst 
rm -rf /var/run/knox 
rm -rf /var/run/oozie 
rm -rf /var/run/webhcat 
rm -rf /var/run/zookeeper 

11 删除所有节点上的库文件夹 

rm -rf /usr/lib/ambari-agent 
rm -rf /usr/lib/ambari-infra-solr-client 
rm -rf /usr/lib/ambari-metrics-hadoop-sink 
rm -rf /usr/lib/ambari-metrics-kafka-sink 
rm -rf /usr/lib/ambari-server-backups 
rm -rf /usr/lib/ams-hbase 
rm -rf /usr/lib/mysql 
rm -rf /var/lib/ambari-agent 
rm -rf /var/lib/ambari-metrics-grafana 
rm -rf /var/lib/ambari-server 
rm -rf /var/lib/flume 
rm -rf /var/lib/hadoop-hdfs 
rm -rf /var/lib/hadoop-mapreduce 
rm -rf /var/lib/hadoop-yarn 
rm -rf /var/lib/hive2 
rm -rf /var/lib/knox 
rm -rf /var/lib/smartsense 
rm -rf /var/lib/storm

12 清除所有节点上的文件夹/var/tmp/*

rm -rf /var/tmp/* 

13 从cron在所有节点上删除HST 

/usr/hdp/share/hst/bin/hst-scheduled-capture.sh sync 
/usr/hdp/share/hst/bin/hst-scheduled-capture.sh

14 删除数据库

删除MySQL和Postgres的实例，以便Ambari安装和配置新的数据库

yum erase postgresql 
rm -rf /var/lib/pgsql 
rm -rf /var/lib/mysql 

15 删除所有节点上的符号链接

检查文件夹/usr/sbin和/usr/lib/python2.6/site-packages

cd /usr/bin 
rm -rf accumulo 
rm -rf atlas-start 
rm -rf atlas-stop 
rm -rf beeline 
rm -rf falcon 
rm -rf flume-ng 
rm -rf hbase 
rm -rf hcat 
rm -rf hdfs 
rm -rf hive 
rm -rf hiveserver2 
rm -rf kafka 
rm -rf mahout 
rm -rf mapred 
rm -rf oozie 
rm -rf oozied.sh 
rm -rf phoenix-psql 
rm -rf phoenix-queryserver 
rm -rf phoenix-sqlline 
rm -rf phoenix-sqlline-thin 
rm -rf pig 
rm -rf python-wrap 
rm -rf ranger-admin 
rm -rf ranger-admin-start 
rm -rf ranger-admin-stop 
rm -rf ranger-kms 
rm -rf ranger-usersync 
rm -rf ranger-usersync-start 
rm -rf ranger-usersync-stop 
rm -rf slider 
rm -rf sqoop 
rm -rf sqoop-codegen 
rm -rf sqoop-create-hive-table 
rm -rf sqoop-eval 
rm -rf sqoop-export 
rm -rf sqoop-help 
rm -rf sqoop-import 
rm -rf sqoop-import-all-tables 
rm -rf sqoop-job 
rm -rf sqoop-list-databases 
rm -rf sqoop-list-tables 
rm -rf sqoop-merge 
rm -rf sqoop-metastore 
rm -rf sqoop-version 
rm -rf storm 
rm -rf storm-slider 
rm -rf worker-lanucher 
rm -rf yarn 
rm -rf zookeeper-client 
rm -rf zookeeper-server 
rm -rf zookeeper-server-cleanup

16 删除所有节点上的服务用户

userdel -r accumulo 
userdel -r ambari-qa 
userdel -r ams 
userdel -r falcon 
userdel -r flume 
userdel -r hbase 
userdel -r hcat 
userdel -r hdfs 
userdel -r hive 
userdel -r kafka 
userdel -r knox 
userdel -r mapred 
userdel -r oozie 
userdel -r ranger 
userdel -r spark 
userdel -r sqoop 
userdel -r storm 
userdel -r tez 
userdel -r yarn 
userdel -r zeppelin 
userdel -r zookeeper 

17 清理遗留文件

在所有节点上运行find / -name **  你一定会找到更多的文件/文件夹。 删除它们

find / -name ambari 
find / -name accumulo 
find / -name atlas 
find / -name beeline 
find / -name falcon 
find / -name flume 
find / -name hadoop 
find / -name hbase 
find / -name hcat 
find / -name hdfs 
find / -name hdp 
find / -name hive 
find / -name hiveserver2 
find / -name kafka 
find / -name mahout 
find / -name mapred 
find / -name oozie 
find / -name phoenix 
find / -name pig 
find / -name ranger 
find / -name slider 
find / -name sqoop 
find / -name storm 
find / -name yarn 
find / -name zookeeper 

18 重新启动所有节点 

reboot