Hi,
I am reading the PR, but I think some design ideas need to be introduced ...
E.g., you change the data file structure from data/storage_group/*.files to
data/storage_group/{time range}/*.files..
Best,
-----------------------------------
Xiangdong Huang
School of Software, Tsinghua University
黄向东
清华大学 软件学院
南京大学软件学院薛恺丰 <[email protected]> 于2019年12月3日周二 上午10:08写道:
> Hi~
> I'm working on this issue. In current storage group there is only one
> working processor for sequence data and one working processor for
> unsequence data. We will use a LRU map to hold working processor in
> different time range (timeRange -> processor). Data in different time
> range will insert into different processor. Last frequent used processor
> will flush and close. If you have some ideas, please feel free to discuss
> with me.
>
>
>
>
> ------------------ 原始邮件 ------------------
> 发件人: "Yuan Tian (Jira)"<[email protected]>;
> 发送时间: 2019年12月3日(星期二) 上午9:52
> 收件人: "dev"<[email protected]>;
>
> 主题: [jira] [Created] (IOTDB-333) Storage group needs to divide by
> time range
>
>
>
> Kaifeng Xue created IOTDB-333:
> ---------------------------------
>
>
> Summary: Storage group needs to divide by time range
>
> Key: IOTDB-333
>
> URL: https://issues.apache.org/jira/browse/IOTDB-333
>  
> <https://issues.apache.org/jira/browse/IOTDB-333  >;
> Project: Apache IoTDB
> Issue Type: New
> Feature
>
> Reporter: Kaifeng Xue
>
>
> Our storage group needs to divide by time range, which can support
> distributed model better.
>
>
>
> --
> This message was sent by Atlassian Jira
> (v8.3.4#803005)