20 分钟自动搭建大数据平台

docker pull cloudera/quickstart
// 或者在 cloudera 官网下载文件后使用 `docker import` 命令导入
// https://downloads.cloudera.com/demo_vm/docker/cloudera-quickstart-vm-5.13.0-0-beta-docker.tar.gz

vim /etc/hosts
xx.xx.xx.xx quickstart.cloudera

docker run -d --name cdh  --hostname=quickstart.cloudera --privileged=true -p 8020:8020 -p 8022:8022 -p 7180:7180 -p 21050:21050 -p 50070:50070 -p 50075:50075 -p 50010:50010 -p 50020:50020 -p 8890:8890 -p 60010:60010 -p 10002:10002 -p 25010:25010 -p 25020:25020 -p 18088:18088 -p 8088:8088 -p 19888:19888 -p 7187:7187 -p 11000:11000 -p 8888:8888 -p 2181:2181 -p 10000:10000 cloudera/quickstart /bin/bash -c '/usr/bin/docker-quickstart && /home/cloudera/cloudera-manager --express --force && service ntpd start && tail -F /var/log/*.log'

user_id,name,sex,age
10001,张三,1,20
10002,李四,0,18
10003,王五,1,27
10004,赵六,1,33

user_id,fans_id,time
10001,10002,2019-10-01
10001,10003,2019-11-03
10002,10003,2019-10-22
10002,10004,2019-11-02
10003,10001,2019-09-13
10004,10001,2019-09-08
10004,10002,2019-10-08
10004,10003,2019-11-15

pip install hdfs

import hdfs

client = hdfs.InsecureClient('http://quickstart.cloudera:50070', user='cloudera')

# 浏览目录
print(client.list('/'))

# 创建目录
client.makedirs('/user/cloudera/users')
client.makedirs('/user/cloudera/fans')

# 上传文件
client.upload('/user/cloudera/users/data.csv', './data1.csv', overwrite=True)
client.upload('/user/cloudera/fans/data.csv', './data2.csv', overwrite=True)
print('upload success!')

# 读取 hdfs 文件内容
with client.read('/user/cloudera/users/data.csv') as r:
    print(r.read())

create external table users
(user_id int, name string, sex int, age int)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY ','
STORED AS TEXTFILE
location '/user/cloudera/users'
tblproperties("skip.header.line.count"="1");

create external table fans
(user_id int, fans_id int, time string)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY ','
STORED AS TEXTFILE
location '/user/cloudera/fans'
tblproperties("skip.header.line.count"="1");

SELECT name, count(*) AS fans_count
FROM users
LEFT JOIN fans
    ON fans.user_id = users.user_id
GROUP BY  name;

apt-get install libsasl2-dev -y
pip install sasl
pip install thrift
pip install thrift-sasl
pip install PyHive

from pyhive import hive
conn = hive.Connection(host='quickstart.cloudera', port=10000, username='cloudera', database='default')
cursor = conn.cursor()

cursor.execute('select * from users;')
print(cursor.fetchall())

cursor.execute('select count(*) from users;')
print(cursor.fetchone())

cursor.close()
conn.close()