Apache Drill POC调研，输出Apache Drill安装部署README.md about alldata

[YueCang]

Drill安装

Drill on docker

docker run -it --name drill \
	-p 8047:8047 \		# web and REST
	-p 31010:31010 \	# JDBC
	apache/drill-314
	
	
Apache Drill 1.20.3
"What ever the mind of man can conceive and believe, Drill can query."
apache drill> select version from sys.version;
+---------+
| version |
+---------+
| 1.20.3  |
+---------+
1 row selected (1.917 seconds)
apache drill>

web-site: <http://<ip>:8047/>

Drill on the cluster

• 安装JDK

• 安装zookeeper

• 安装&配置

  • 解压

    tar -zxvf apache-drill-1.20.0.tar.gz -C /opt/
    
    mv apache-drill-1.20.0 drill-1.20.0

  • 配置

    mdkir /opt/drill-1.20.0/drill-site
    
    cp /opt/drill-1.20.0/conf/drill-override.conf /opt/drill-1.20.0/drill-site
    
    mkdir /opt/drill-1.20.0/drill-site/data
    
    vim drill-site
    drill.exec: {
      cluster-id: "drill-cluster",
      zk.connect: "hadoop102:2181,hadoop103:2181,hadoop104:2181",
      sys.store.provider.local.path="/opt/drill-1.20.0/drill-site/data"
    }

    注:集群每一台节点都需要安装&配置

• 启动&测试

  • 启动zookeeper

  • 启动

    # 每台节点都需要启动
    ./drillbit.sh --config ../drill-site/ start

  • 测试

    bin/sqlline -u jdbc:drill:zk=hadoop102,hadoop103,hadoop104:2181/drill/<cluster-id>
    
    Apache Drill 1.20.0
    "Drill, baby, Drill."
    apache drill> select version from sys.version;
    +---------+
    | version |
    +---------+
    | 1.20.0  |
    +---------+
    1 row selected (0.192 seconds)
    apache drill> select * from dfs.`/opt/drill-1.20.0/drill-site/data/test.csv`;
    +--------------+
    |   columns    |
    +--------------+
    | ["id\tname"] |
    | ["1\tlyb"]   |
    | ["2\tley"]   |
    +--------------+
    3 rows selected (0.187 seconds)

[alldatafounder]

Instead of just running Drill in docker, we need the process of compiling, deploying and installing the source code to a designated server, outputting process documents, and collecting and sorting out ideas related to integration with studio

[alldatafounder]

not plan