jackrabbit-oak-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jörg Hoh (JIRA) <j...@apache.org>
Subject [jira] [Commented] (OAK-3141) [DocumentNodeStore] Oak should warn when large documents are written
Date Mon, 03 Aug 2015 13:46:04 GMT

    [ https://issues.apache.org/jira/browse/OAK-3141?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14651867#comment-14651867

Jörg Hoh commented on OAK-3141:

[~julian.reschke@greenbytes.de] Yes, I think, that's the better approach: WARN, when a node
as excessive number of ordered children. I would suggest, that 2000 might be a good value
(of course configurable).

> [DocumentNodeStore] Oak should warn when large documents are written
> --------------------------------------------------------------------
>                 Key: OAK-3141
>                 URL: https://issues.apache.org/jira/browse/OAK-3141
>             Project: Jackrabbit Oak
>          Issue Type: Improvement
>          Components: core
>    Affects Versions: 1.0.16
>            Reporter: Jörg Hoh
> When working with the RDBMK we came into situations, that large documents did not fit
into the provided db columns, there was an overflow, which caused oak not to persist the change.
We fixed it by increasing the size of the column.
> But it would be nice if Oak could warn if a document exceeds a certain size (for example
2 megabytes); because this warning indicates, that on a JCR level there might be a problematic
situation, for example:
> * ordered node with a large list of childnodes
> * or longstanding sessions with lots of changes, which accumulate to large documents.
> It's certainly nice to know if there's a node/document with such a problem, before the
exceptions actually happens and an operation breaks.
> This message should be a warning, and should contain the JCR path of the node plus the
current size. To avoid that this message is overseen, it would be good if it is written everyonce
in a while (every 10 minutes?) if this condition persists.

This message was sent by Atlassian JIRA

View raw message