[jira] [Commented] (OAK-6351) Invalidate cache entries when getChildNodes() is aborted

["Julian Reschke (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/OAK-6351?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16064731#comment-16064731 ] 

Julian Reschke commented on OAK-6351:
-------------------------------------

And the aforementioned NPE indeed happened on an 1.0.* system:

{noformat}
java.lang.NullPointerException: null
at org.apache.jackrabbit.oak.plugins.document.DocumentNodeStore$9.apply(DocumentNodeStore.java:1068)
at org.apache.jackrabbit.oak.plugins.document.DocumentNodeStore$9.apply(DocumentNodeStore.java:1057)
at com.google.common.collect.Iterators$8.transform(Iterators.java:794)
{noformat}

...so this should be backported.

> Invalidate cache entries when getChildNodes() is aborted
> --------------------------------------------------------
>
>                 Key: OAK-6351
>                 URL: https://issues.apache.org/jira/browse/OAK-6351
>             Project: Jackrabbit Oak
>          Issue Type: Improvement
>          Components: documentmk
>            Reporter: Marcel Reutegger
>            Assignee: Marcel Reutegger
>            Priority: Minor
>              Labels: candidate_oak_1_0, candidate_oak_1_2, candidate_oak_1_4, candidate_oak_1_6
>             Fix For: 1.8, 1.7.2
>
>         Attachments: OAK-6351.patch
>
>
> With OAK-4623 {{DocumentNodeStore.getChildNodes()}} logs the document of the missing child node read directly from the document store. This effectively also ensures the document cache now contains an up-to-date version of the document.
> In addition the relevant cache entries in the node state and children caches should also be invalidated. As reported in OAK-6294, the system may still be stuck in this aborting state even if the document is now up-to-date.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)