[ 
https://issues.apache.org/jira/browse/OOZIE-1035?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13677259#comment-13677259
 ] 

Robert Kanter commented on OOZIE-1035:
--------------------------------------

Your first example definitely looks like something we should be allowing; the 
second one I'd have to think about some more.  

It's been a while since I looked at this code, so I'll need to spend some time 
going through it again to look into this.  

In the mean time, you can disable the forkjoin checking either for every 
workflow (by setting {{oozie.validate.ForkJoin}} to {{false}} in 
oozie-site.xml) or just for a specific workflow (by setting 
{{oozie.wf.validate.ForkJoin}} to {{false}} in your job.properties for that 
workflow).  In light of these flags, it seems better to err on the side of 
caution and make the checking a little bit more restrictive to prevent loops 
even if it's blocking some valid edge cases; this will protect users who are 
beginners but still allow power users to disable the checking and do more 
advanced things.  
                
> Improve forkjoin validation to allow same errorTo transitions
> -------------------------------------------------------------
>
>                 Key: OOZIE-1035
>                 URL: https://issues.apache.org/jira/browse/OOZIE-1035
>             Project: Oozie
>          Issue Type: Improvement
>          Components: workflow
>    Affects Versions: trunk
>            Reporter: Robert Kanter
>            Assignee: Robert Kanter
>            Priority: Minor
>             Fix For: trunk, 3.3.2
>
>         Attachments: OOZIE-1035.patch, OOZIE-1035.patch
>
>
> It seems common that users will have the "error to" transition from every 
> action go to the same action node (e.g. email action), which then goes to the 
> kill node instead of just going to the kill node directly.  When this is done 
> in action nodes within a forkjoin path, the forkjoin validation doesn't allow 
> it.  We should improve the forkjoin validation code to allow the same "error 
> to" transition, as long as it eventually leads to a kill node.  

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to