Docker實戰--(一)

1、在Docker容器中運行Hadoop集羣

3個節點的hadoop集羣搭建

1. pull docker image

[AppleScript] 純文本查看 複製代碼
?
1
sudo docker pull kiwenlau/hadoop:1.0

1. clone github repository

[AppleScript] 純文本查看 複製代碼
?
1
git clone [url=https://github.com/kiwenlau/h...]https://github.com/kiwenlau/hadoop-cluster-docker[/url]

1. create hadoop network

[AppleScript] 純文本查看 複製代碼
?
1
sudo docker network create --driver=bridge hadoop

1. start container

[AppleScript] 純文本查看 複製代碼
?
1
2
cd hadoop-cluster-docker
sudo ./start-container.sh

輸出:

start hadoop-master container...
start hadoop-slave1 container...

start hadoop-slave2 container...

root@hadoop-master:~#

啓動3個容器，包含1個主節點和2個從節點
你將進入hadoop-master容器的/ root目錄

1. start hadoop

[AppleScript] 純文本查看 複製代碼
?
1
./start-hadoop.sh

1. run wordcount

[AppleScript] 純文本查看 複製代碼
?
1
./run-wordcount.sh

輸出

input file1.txt:

Hello Hadoop

input file2.txt:
Hello Docker

wordcount output:

Docker 1

Hadoop 1

Hello 2
二、任意大小的Hadoop集羣

2.1. rebuild docker image

[AppleScript] 純文本查看 複製代碼
?
1
sudo ./resize-cluster.sh 5

指定參數> 1：2,3 ..

這個腳本只是使用不一樣的從屬文件重建hadoop映像，這樣能夠強制全部從屬節點的名稱

2.1. start container[AppleScript] 純文本查看 複製代碼?1sudo ./start-container.sh 5