ambari-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Varun Kapoor (JIRA)" <>
Subject [jira] [Updated] (AMBARI-205) Revamp The Disk Mount Points Page
Date Thu, 10 May 2012 01:24:49 GMT


Varun Kapoor updated AMBARI-205:

    Remaining Estimate: 12h  (was: 48h)
     Original Estimate: 12h  (was: 48h)
> Revamp The Disk Mount Points Page
> ---------------------------------
>                 Key: AMBARI-205
>                 URL:
>             Project: Ambari
>          Issue Type: Bug
>    Affects Versions: ambari-186
>            Reporter: Varun Kapoor
>             Fix For: ambari-186
>   Original Estimate: 12h
>  Remaining Estimate: 12h
> New Proposal
> * Remove the left panel on that page that provides global (de)selection support.
>    * Every component that makes use of disk mount points will have its own individual
sections, each with its own edit options.
> * Add a global input field that takes in the log+pid directory prefix.
>    * Starts off hidden and expands on click because this is typically an advanced option.
>       * And also because this is a UI anomaly compared to the rest of the page.
> * Add 2 tall columns in place of the one currently on the right.
>    * The column on the left contains the NameNode, JobTracker, Oozie and ZooKeeper (i.e.
Master) data directories.
>    * The column on the right contains the DataNode and TaskTracker (i.e. Slave) data
>       * Each of these sections contains a small "plus" button at the bottom, to add custom
mount points, one per line.
>    * Each of the sections for the NN, JT, DN and TT contain a list of data directories,
with the mount point prefix highlighted in bold (contrasted with the per-component suffix
that'll be displayed in a lighter tone).
>       * Every element in each list will feature a "plus/minus" button that appears only
on hover, that allows people to remove/re-add these individual elements.
>    * The Oozie and ZooKeeper data directories will be rendered as dropdowns, because
only one data directory needs to be selected for these.
>       * Entering custom directories here might be a bit messy, and needs to be mocked
> * Instead of arbitrarily using the NameNode as a canary from which to discover the available
mount points, every component's representative mount point information will be sent to the
>    * Thus, each section shows data directories based off of coherent, relevant mount

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:!default.jspa
For more information on JIRA, see:


View raw message