[jira] [Commented] (OAK-1451) Expose index size

["Alex Parvulescu (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/OAK-1451?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13949041#comment-13949041 ] 

Alex Parvulescu commented on OAK-1451:
--------------------------------------

There is one problem, we don't have this information yet.

The issue should reflect the fact that we need to come up with a solution to persisting the index size first. Some work has been started by [~tmueller] here, but it only went so far as initializing the property index count value. The background thread that was supposed to keep it up to date asynchronously is still not implemented.
So what we currently do is try to extrapolate the property index cost (basically the size) based on the number of key-value reads on the index itself [0].

Thomas, did I get it right?


[0] https://github.com/apache/jackrabbit-oak/blob/trunk/oak-core/src/main/java/org/apache/jackrabbit/oak/plugins/index/property/strategy/ContentMirrorStoreStrategy.java#L182

> Expose index size
> -----------------
>
>                 Key: OAK-1451
>                 URL: https://issues.apache.org/jira/browse/OAK-1451
>             Project: Jackrabbit Oak
>          Issue Type: Improvement
>          Components: query
>            Reporter: Michael Marth
>            Assignee: Alex Parvulescu
>            Priority: Minor
>              Labels: production, resilience
>             Fix For: 0.20
>
>
> At the moment some MK's disc needs are largely from the indexes. Maybe we can do something about this, but in the meantime it would be helpful if we could expose the index sizes (num of indexed nodes) via JMX so that they could be easily monitored.
> This would also be helpful to see at which point an index becomes useless (if the majority of content nodes are indexed one might as well not have an index)



--
This message was sent by Atlassian JIRA
(v6.2#6252)