Rerun oozie workflows every day - hadoop

Question

I have an Oozie coordinator that when started sets its start time to 365 days ago and then runs its workflow with a daily frequency until reaching the latest date having input data. So the coordinator.xml looks something like this:
<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<coordinator-app name="${jobName}" frequency="${coord:days(1)}" start="${startTime}" end="${endTime}" timezone="${timezone}" xmlns="uri:oozie:coordinator:0.2">
<controls>
<timeout>-1</timeout>
<concurrency>30</concurrency>
<execution>FIFO</execution>
</controls>
<datasets>
<dataset name="myData" frequency="${coord:days(1)}" initial-instance="${initialTime}" timezone="${timezone}">
<uri-template>${outputPath}/${YEAR}${MONTH}${DAY}</uri-template>
</dataset>
</datasets>
<input-events>
<data-in name="myInput" dataset="myData">
<instance>${coord:current(-1)}</instance>
</data-in>
</input-events>
<action>
<workflow>
<app-path>${appPath}</app-path>
<configuration>
<property>
<name>myInput</name>
<value>${coord:dataIn('myInput')}</value>
</property>
<property>
<name>date</name>
<value>${coord:formatTime(coord:dateOffset(coord:nominalTime(), -1, 'DAY'), 'yyyyMMdd')}</value>
</property>
</configuration>
</workflow>
</action>
</coordinator-app>
On startup, this coordinator triggers 365 workflows. Then the next day, it just runs one workflow for that day's input data. And so on for the next day.
Now the requirements have changed and I need to run 365 workflows every day, not just the first time! So on startup, I want it to run like it does now. But then the next day, when the input data becomes available, I need to go back and run daily workflows for the last 365 days again. And so on for each subsequent day as soon as new input data is detected.
How can I achieve this?
Thanks,
Alvaro

Answer 1

In the data-in tag of the coordinator use start-instance and end-instance to specify the range of instance to consider as input.
start-instance: refers to the beginning of an instance range (the time for a synchronous dataset).
end-instance: refers to the end of an instance range (the time for a synchronous dataset).
<input-events>
<data-in name="myInput" dataset="myData">
<start-instance>${coord:current(-365)}</start-instance>
<end-instance>${coord:current(-1)}</end-instance>
</data-in>
</input-events>
For more information please refer: Synchronous Coordinator Application Definition

Answer 2

You can use the sub workflow to recall the main workflow in recursive manner.