[
https://issues.apache.org/jira/browse/HAWQ-301?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15080652#comment-15080652
]
ASF GitHub Bot commented on HAWQ-301:
-------------------------------------
GitHub user jiny2 opened a pull request:
https://github.com/apache/incubator-hawq/pull/238
HAWQ-301. Use combined value to measure segment workload to improve t…
In one segment, the combined workload counts:
1) total number of virtual segments allocated;
2) resource usage rate ( used/allocated );
3) estimated accumulated IO workload;
The new vsegs are allocated in the segments having lower combined workload.
If data loclaity information is provided when one client requests query
resource, the unbalance of number of vsegs among segments are restricted.
double rm_regularize_io_max; // maximum limit of accumulated estimated
io workload in one segment for workload regularization.
double rm_regularize_nvseg_max; // maximum limit of accumulated number
of vsegs in one segment for workload regularization.
double rm_regularize_io_factor; // factor of accumulated estimated io
workload, 1 as default.
double rm_regularize_usage_factor; // factor of resource usage, 1 as
default.
double rm_regularize_nvseg_factor; // factor of accumulated number of
vsegs, 1 as default.
int rm_nvseg_variance_among_seg_respool_limit; // the limit of
unbalance of accumulated number of vsegs among segments after getting tentative
vseg allocation based on data locality information.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/jiny2/incubator-hawq HAWQ-301
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/incubator-hawq/pull/238.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #238
----
commit 58306357931655b9c486c5af460ef5a5e540a0af
Author: YI JIN <[email protected]>
Date: 2016-01-04T03:03:21Z
HAWQ-301. Use combined value to measure segment workload to improve the
resource allocation strategy
----
> Use combined value to measure segment workload to improve the resource
> allocation strategy
> ------------------------------------------------------------------------------------------
>
> Key: HAWQ-301
> URL: https://issues.apache.org/jira/browse/HAWQ-301
> Project: Apache HAWQ
> Issue Type: Improvement
> Components: Resource Manager
> Reporter: Yi Jin
> Assignee: Yi Jin
>
> In one segment, the combined workload counts:
> 1) total number of virtual segments allocated;
> 2) resource usage rate ( used/allocated );
> 3) estimated accumulated IO workload;
> The new vsegs are allocated in the segments having lower combined workload.
> If data loclaity information is provided when one client requests query
> resource, the unbalance of number of vsegs among segments are restricted.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
