[
https://issues.apache.org/jira/browse/TAJO-1703?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14639907#comment-14639907
]
Jinho Kim commented on TAJO-1703:
---------------------------------
Here is my test result.
* 1 master + 6 worker
** CPU: 48 core, Memory: 128GB, Disk: SAS 7.2K * 12
* TPC-H scale 1TB
|Status update rate| 50%| 30%|
|select count(*) from lineitem| 110 sec| 102 sec|
|Master CPU usage| 0.3| 0.7|
|Locality| 99.984| 99.984|
> Remove hardcoded value in NodeStatusUpdater
> -------------------------------------------
>
> Key: TAJO-1703
> URL: https://issues.apache.org/jira/browse/TAJO-1703
> Project: Tajo
> Issue Type: Improvement
> Components: Resource Manager, Worker
> Reporter: Jinho Kim
> Assignee: Jinho Kim
> Fix For: 0.11.0
>
>
> NodeStatusUpdater updates the current status when resources is changed and it
> determined by threshold.
> If a server is very high spec hardware, lower value is better in order to
> low latency.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
