[
https://issues.apache.org/jira/browse/KYLIN-2265?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15739885#comment-15739885
]
Billy Liu commented on KYLIN-2265:
----------------------------------
Hi Shaofeng, do you have more suggestions on what the default would be. I try
to propose the sandbox(10G memory, 2 core) as default, because it has the least
limitations to the hadoop environment. The default configuration should build
sample cube smoothly. For production deployment, the administrator should
reconsider the cluster size, and source data size carefully. The configuration
document is mandatory for reference.
> Adjust default config for resource-limit sandbox
> ------------------------------------------------
>
> Key: KYLIN-2265
> URL: https://issues.apache.org/jira/browse/KYLIN-2265
> Project: Kylin
> Issue Type: Improvement
> Components: General
> Affects Versions: v1.6.0
> Reporter: Billy Liu
> Assignee: Billy Liu
> Attachments: KYLIN-2265.patch
>
>
> Most users deploy Kylin on their own sandbox for the first trial. Most
> sandbox has most 8G memory. The most used sandboxes are HDP sandbox and CDH
> sandbox. We'd better make the default kylin configuration convenient for
> these sandbox environment. The suggestion includes reducing region-cut-gb,
> hfile-size-gb, max-region-count, reduce-input-mb,
> max-reducer-number,mapreduce.map.memory.mb,mapreduce.map.java.opts.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
