[ 
https://issues.apache.org/jira/browse/ATLAS-2807?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16568010#comment-16568010
 ] 

Graham Wallis commented on ATLAS-2807:
--------------------------------------

Hi [[email protected]]

Thanks for providing the PDF which explains what this Jira is about. I have a 
question: does the desired behaviour depend on the nature of the relationship? 
I can see that where an entity has been created (t1 in your examples) then it 
probably should retain the tags pertaining to the entity and process from which 
it was created (even if the source of that lineage is subsequently deleted). 
But for other types of relationship it seems less obvious to me that we would 
want to retain the tags from the 'upstream' entity if it is deleted.

> Re-evaluate classification propagation during entity delete
> -----------------------------------------------------------
>
>                 Key: ATLAS-2807
>                 URL: https://issues.apache.org/jira/browse/ATLAS-2807
>             Project: Atlas
>          Issue Type: Improvement
>          Components:  atlas-core
>    Affects Versions: 1.0.0-alpha, 1.0.0
>            Reporter: Sarath Subramanian
>            Assignee: Sarath Subramanian
>            Priority: Major
>             Fix For: 1.1.0
>
>         Attachments: 
> 0001-ATLAS-2807-Re-evaluate-classification-propagation-du.patch, 
> Classification Propagation during entity delete1.pdf
>
>
> The current behavior - when we delete an entity *f1*, all tags associate to 
> *f1* which got propagated
> to downstream entities are removed – *PII* tag propagated to *process1* and 
> *t1* is removed
> *Proposed Change*
> _*Soft Delete:*_
> When entity *f1* is deleted, retain the propagated classification edges, so 
> the downstream entities
> – *process1* and *t1* continue to have *PII* classification associated to them
> _*Hard Delete:*_
> In cases of hard delete, the source entity (*f1*) is deleted from atlas, but 
> its classification vertex
> (*PII*) continue to exist and will continue propagating to *process1* and 
> *t1*.
>  



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to