accumulo-notifications mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "John Vines (JIRA)" <j...@apache.org>
Subject [jira] [Comment Edited] (ACCUMULO-780) Accumulo should have a configurator
Date Mon, 24 Feb 2014 20:34:29 GMT

    [ https://issues.apache.org/jira/browse/ACCUMULO-780?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13910774#comment-13910774
] 

John Vines edited comment on ACCUMULO-780 at 2/24/14 8:33 PM:
--------------------------------------------------------------

It doesn't seem to distinguish that the 3GB configs are for a cluster (individual processes
will take 3GB max as opposed to all processes together taking up that footprint).

Furthermore, my original intention for this ticket was to do some heuristics so users could
specify different values to generate a list of settings instead of being stuck with 4 different
size options.


was (Author: vines):
It doesn't seem to distinguish that the 3GB configs are for a cluster (individual processes
will take 3GB max as opposed to all processes together taking up that footprint).

Furthermore, my original intention for this ticket, which I did not convey, was to do some
heuristics so users could specify different values to generate a list of settings instead
of being stuck with 4 different size options.

> Accumulo should have a configurator
> -----------------------------------
>
>                 Key: ACCUMULO-780
>                 URL: https://issues.apache.org/jira/browse/ACCUMULO-780
>             Project: Accumulo
>          Issue Type: New Feature
>          Components: scripts
>            Reporter: John Vines
>            Assignee: John McNamee
>             Fix For: 1.7.0
>
>         Attachments: ACCUMULO-780.v1.patch
>
>
> We currently have a few different footprints available for users. We should use those
as a base and allow users to quickly and conveniently configure their system for whatever
size footprint they want. We can use the current as a baseline and extrapolate/interpolate
for whichever setup they're using. This way we don't have to worry about maintaining a bunch
of different sizes and instead just have an algorithm that needs occasional loving.



--
This message was sent by Atlassian JIRA
(v6.1.5#6160)

Mime
View raw message