[jira] [Updated] (HBASE-15640) L1 cache doesn't give fair warning that it is showing partial stats only when it hits limit

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

     [ https://issues.apache.org/jira/browse/HBASE-15640?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

stack updated HBASE-15640:
--------------------------
    Attachment: bc.ui.patch
                Screen Shot 2016-04-12 at 4.21.50 PM.png

Here is small patch which ups our limit to 1M blocks from 100k so less likely we'll hit limit and then if we do, puts in place a message in red with how to address the limit (I couldn't figure what the previous message was trying to say... didn't seem to make sense).

This could be better.. but will do for now.

> L1 cache doesn't give fair warning that it is showing partial stats only when it hits limit
> -------------------------------------------------------------------------------------------
>
>                 Key: HBASE-15640
>                 URL: https://issues.apache.org/jira/browse/HBASE-15640
>             Project: HBase
>          Issue Type: Sub-task
>          Components: BlockCache
>            Reporter: stack
>            Assignee: stack
>         Attachments: Screen Shot 2016-04-12 at 4.21.50 PM.png, bc.ui.patch
>
>
> I was baffled looking at L1 Cache because it was stuck at 100k blocks and only 'loading' 3G of data according to the UI. When I looked in log, all the numbers looked write in the Cache summary reported in log.
> Turns out our buckecache UI template will cut off calculating stats at a limit --which is probably what we want -- but the fact that it has done this should be more plain.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)