[
https://issues.apache.org/jira/browse/ATLAS-3545?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16985659#comment-16985659
]
Bolke de Bruin commented on ATLAS-3545:
---------------------------------------
Stack trace is unavailable at the moment as Atlas swallows the exception. We
will patch it to make sure the whole trace is available.
We are observing 3 separate issues that are related. 1) we set a classification
on a aws_s3_bucket with several thousands (200K+) of objects. As the UI wasn’t
clear if the classification was applied so we tried it a couple of times. This
resulted in the UI displaying it 3 (!) times. We then tried to remove one,
which seemed to work but trying to remove the others fails.
Moreover, the original classification still seems to be there as we face
scaleability issues when updating the aws_s3_bucket in question: it then
eagerly fetches all vertices that have a propagating tag which exponentially
increases the time spend on this when s3_objects are added. This will be filed
separately in a jira.
> NullPointerException while trying to delete classification
> ----------------------------------------------------------
>
> Key: ATLAS-3545
> URL: https://issues.apache.org/jira/browse/ATLAS-3545
> Project: Atlas
> Issue Type: Bug
> Affects Versions: trunk
> Reporter: Bolke de Bruin
> Priority: Critical
> Attachments: Screenshot 2019-11-30 at 22.16.44.png, Screenshot
> 2019-11-30 at 22.28.00.png
>
>
> We see an issue where there is a NullPointerException while trying to delete
> a classification that is propagating.
> It seems (stack trace is as of yet unavailable) that it is caused due to a
> typeName being NULL in AtlasTypeRegistry.getType. The UI looks fishy too see
> screenshots.
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
