You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@kylin.apache.org by "ShengJun Zheng (Jira)" <ji...@apache.org> on 2021/03/12 02:14:00 UTC

[jira] [Updated] (KYLIN-4933) Support set cache strength for dict cache

     [ https://issues.apache.org/jira/browse/KYLIN-4933?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

ShengJun Zheng updated KYLIN-4933:
----------------------------------
    Affects Version/s: v3.1.1

> Support set cache strength for dict cache
> -----------------------------------------
>
>                 Key: KYLIN-4933
>                 URL: https://issues.apache.org/jira/browse/KYLIN-4933
>             Project: Kylin
>          Issue Type: Improvement
>    Affects Versions: v3.1.1
>            Reporter: ShengJun Zheng
>            Priority: Major
>
> Kylin uses guava LoadingCache(LUR Cache) to cache dictionary with soft(SoftReference) cache strength , it will be collected due to JVM GC. 
> In cases of queries over  dict-encoded-dimension cubes, we would always prefer to cache dimension dictionary in LUR cache to avoid reloading dict from HBase/HDFS, which will causes unstable query performance.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)