Azkaban combat

Azkaba built-in task types support command, java

Command type single job example

1, create a job description file

vi command.job

command.job

type = command

command = echo 'hello'

2, the job resource files packaged into a zip file

zip command.job

3, through the azkaban web management platform to create a project and upload the job archive

First create the project

Upload zip package

4, start the implementation of the job

Command type multiple job workflow flow

1, create a dependency of multiple job description

The first job: foo.job

foo.job

type = command

command = echo foo

The second job: bar.job depends on foo.job

bar.job

type = command

dependencies = foo

command = echo bar

2, all job resource files hit a zip package

3, create a project in azkaban web management interface and upload zip package

4, start the workflow flow

HDFS operation task

1, create a job description file

fs.job

type = command

command = / home / hadoop / apps / hadoop-2.6.1 / bin / hadoop fs -mkdir / azaz

2, the job resource files packaged into a zip file

3, through the azkaban web management platform to create a project and upload the job archive

4, start the implementation of the job

MAPREDUCE task

Mr task still can use the command's job type to execute

1, create a job description file, and mr program jar package (example directly use hadoop own example jar)

mrwc.job

type = command

command = / home / hadoop / apps / hadoop-2.6.1 / bin / hadoop jar hadoop-mapreduce-examples-2.6.1.jar wordcount / wordcount / input / wordcount / azout

2, all job resource files hit a zip package

3, create a project in azkaban web management interface and upload zip package

4, start the job

HIVE script task

Create a job description file and hive script

Hive script: test.sql

use default;

drop table aztest;

create table aztest (id int, name string) row format delimited fields terminated by ',';

load data inpath '/ aztest / hiveinput' into table aztest;

create table azres as select * from aztest;

insert overwrite directory '/ aztest / hiveoutput' select count (1) from aztest;

Job description file: hivef.job

hivef.job

type = command

command = / home / hadoop / apps / hive / bin / hive -f 'test.sql'

2, all job resource files hit a zip package

3, create a project in azkaban web management interface and upload zip package

4, start the job