[ 
https://issues.apache.org/jira/browse/IOTDB-1117?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17296784#comment-17296784
 ] 

Tian Jiang commented on IOTDB-1117:
-----------------------------------

Thanks for your participation, but I guess this is a non-trivial work and 
require sufficient understanding of both the cluster module and the standalone 
query module, so please make sure you have designed for it and discuss your 
design with us before proceeding to development.


在2021年3月6日 18:07,Chao Wang (Jira)<[email protected]> 写道:

[ 
https://issues.apache.org/jira/browse/IOTDB-1117?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17296498#comment-17296498
 ]

Chao Wang 在 IOTDB-1117中留言:
----------------------------------

I want to do it.. 

[Distributed] Batched creation and fetch of RemoteSeriesReader
--------------------------------------------------------------

关键字: IOTDB-1117
URL: https://issues.apache.org/jira/browse/IOTDB-1117
项目: Apache IoTDB
问题类型: 改进
模块: Cluster
报告人: Tian Jiang
优先级: 重要
标签: cluster, query

Each creation and fetch of a RemoteSeriesReader requires one RPC and when there 
are many series involved in a query, the RPC overhead may increase 
significantly and slow down the whole query.
So it is reasonable to group RemoteSeriesReaders by the DataGroup they belong 
to, and send RPCs to create or fetch them in a batched manner.



--
这条信息是由Atlassian Jira发送的
(v8.3.4#803005)


> [Distributed] Batched creation and fetch of RemoteSeriesReader
> --------------------------------------------------------------
>
>                 Key: IOTDB-1117
>                 URL: https://issues.apache.org/jira/browse/IOTDB-1117
>             Project: Apache IoTDB
>          Issue Type: Improvement
>          Components: Cluster
>            Reporter: Tian Jiang
>            Assignee: Chao Wang
>            Priority: Major
>              Labels: cluster, query
>
> Each creation and fetch of a RemoteSeriesReader requires one RPC and when 
> there are many series involved in a query, the RPC overhead may increase 
> significantly and slow down the whole query. 
> So it is reasonable to group RemoteSeriesReaders by the DataGroup they belong 
> to, and send RPCs to create or fetch them in a batched manner.



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

Reply via email to