Apache Druid入门系列(二): 安装配置

Ａpacahe Druid官方一直在持续更新版本，查阅最新版本请参考：http://druid.io/docs/latest/tutorials/index.html
本文是以0.16.0版本为例，说明druid单机版的安装配置过程，参考自：https://druid.apache.org/docs/0.16.0-incubating/tutorials/index.html

在本文示例中，我们将下载druid并将其配置在一台机器上，本例中此机器为虚拟机，IP：192.168.1.164，系统：Centos6.9

环境需求

• Java 8(8u92+)或更高
• Linux, Mac OS X, or other Unix-like OS (不支持Windows系统)
• 8G of RAM
• 2 vCPUs

下载并解压

下载软件包 0.16.0-incubating release.
http://mirrors.tuna.tsinghua.edu.cn/apache/incubator/druid/0.16.0-incubating/apache-druid-0.16.0-incubating-bin.tar.gz

我们把druid安装到 /usr/druid下

cd /usr
mkdir druid
cd druid
# 下载文件包（也可以先下载电脑上，再复制到虚拟机上）
curl http://mirrors.tuna.tsinghua.edu.cn/apache/incubator/druid/0.16.0-incubating/apache-druid-0.16.0-incubating-bin.tar.gz
# 解压文件包：
tar -xzf apache-druid-0.16.0-incubating-bin.tar.gz

你将在目录下发现以下的内容:
$\color{orange}{免责声明、许可和通知文件}$
$\color{orange}{LICENSE}$ - license许可文件.
$\color{orange}{bin/* }$- 非常有用的快速入门脚本文件
$\color{orange}{conf/*}$ - template configurations for a clustered setup
$\color{orange}{extensions/* }$- core Druid extensions
$\color{orange}{hadoop-dependencies/*}$ - Druid Hadoop dependencies
$\color{orange}{lib/* }$- libraries and dependencies for core Druid
$\color{orange}{quickstart/*}$ - configuration files, sample data, and other files for the quickstart tutorials

下载 Zookeeper, 解压 并启动

cd apache-druid-0.16.0-incubating
curl https://archive.apache.org/dist/zookeeper/zookeeper-3.4.14/zookeeper-3.4.14.tar.gz -o zookeeper-3.4.14.tar.gz
tar -xzf zookeeper-3.4.14.tar.gz
mv zookeeper-3.4.14 zk

注：$\color{red}{要保证zk在目录 apache-druid-0.16.0-incubating的根目录下}$

启动Druid服务

在 apache-druid-0.16.0-incubating目录下，执行如下命令

./bin/start-micro-quickstart

也可以在druid目录，写成批处理脚本文件

[root@master-node druid]# cd /usr/druid
[root@master-node druid]# cat start0.16.0.sh
echo 开始启动......
cd apache-druid-0.16.0-incubating
./bin/start-micro-quickstart

[root@master-node druid]# sh cat start0.16.0.sh

你将看到如下的结果:

[Mon Apr 29 00:20:49 2019] Running command[zk], logging to[/usr/druid/apache-druid-0.14.0-incubating/var/sv/zk.log]: bin/run-zk quickstart/tutorial/conf
[Mon Apr 29 00:20:49 2019] Running command[coordinator], logging to[/usr/druid/apache-druid-0.14.0-incubating/var/sv/coordinator.log]: bin/run-druid coordinator quickstart/tutorial/conf
[Mon Apr 29 00:20:49 2019] Running command[broker], logging to[/usr/druid/apache-druid-0.14.0-incubating/var/sv/broker.log]: bin/run-druid broker quickstart/tutorial/conf
[Mon Apr 29 00:20:49 2019] Running command[router], logging to[/usr/druid/apache-druid-0.14.0-incubating/var/sv/router.log]: bin/run-druid router quickstart/tutorial/conf
[Mon Apr 29 00:20:49 2019] Running command[historical], logging to[/usr/druid/apache-druid-0.14.0-incubating/var/sv/historical.log]: bin/run-druid historical quickstart/tutorial/conf
[Mon Apr 29 00:20:49 2019] Running command[overlord], logging to[/usr/druid/apache-druid-0.14.0-incubating/var/sv/overlord.log]: bin/run-druid overlord quickstart/tutorial/conf
[Mon Apr 29 00:20:49 2019] Running command[middleManager], logging to[/usr/druid/apache-druid-0.14.0-incubating/var/sv/middleManager.log]: bin/run-druid middleManager quickstart/tutorial/conf

• 所有持久状态，如集群元数据存储和服务段 都保存到apache-druid-0.16.0-incubating的var目录下. 日志信息保存到 var/sv下.
• 按 CTRL-C 可以退出服务

重置集群状态
如果你希望在停止服务后，下次以初始化状态启动服务，就删除log和var目录

重置kafka
假如你从kafka流加载数据，重启服务还清除kafka态，在停止Zookeeper和Druid服务之前，使用CTRL-C关闭Kafka broker，然后删除/tmp/ Kafka -logs上的Kafka日志目录

rm -rf /tmp/kafka-logs

一旦集群启动，就可以导航到http://192.168.1.164:8888。服务于Druid控制台的Druid路由器进程驻留在这个地址。