flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "sunjincheng (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (FLINK-5657) Add processing time OVER RANGE BETWEEN UNBOUNDED PRECEDING aggregation to SQL
Date Mon, 13 Feb 2017 07:58:41 GMT

    [ https://issues.apache.org/jira/browse/FLINK-5657?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15863303#comment-15863303
] 

sunjincheng commented on FLINK-5657:
------------------------------------

Hi,guys,I made a preliminary implementation of this JIRA.
My approach is:
1. Calcite -> Flink
    "LogicalProject with RexOver expression" --(normalize rule)-> "Calcite's LogicalWindow"
--(opt rule) -> DataStreamRowWindowAggregate

2. datastreamAPI:
  a. With partitionBy situation: 
     approach1: inputDS.map().keyby().reduce().map() 
     approach2: inputDS.map().Keyby().process()
  b. Without paritionBy situation: 
     inputDS.map().setParallelism(1), map has implement CheckPointedFunction.

 3. About OrderBy:
    According to the natural order of elements, procTime () use for generate end-time of the
window and guaranteed  pass the sql validation.

HI,[~fhueske] IMO. “Calcite -> FLINK” part should be rowWindow related JIRAs shared
part, in order to share as soon as possible, I would like to change JIRA. into two subtasks:
  1. rowWindow with partitionBy 
  2. rowWindow without partitionBy.

 It's that make sense for you? I would be very grateful if you could give me some advices.


> Add processing time OVER RANGE BETWEEN UNBOUNDED PRECEDING aggregation to SQL
> -----------------------------------------------------------------------------
>
>                 Key: FLINK-5657
>                 URL: https://issues.apache.org/jira/browse/FLINK-5657
>             Project: Flink
>          Issue Type: Sub-task
>          Components: Table API & SQL
>            Reporter: Fabian Hueske
>            Assignee: sunjincheng
>
> The goal of this issue is to add support for OVER RANGE aggregations on processing time
streams to the SQL interface.
> Queries similar to the following should be supported:
> {code}
> SELECT 
>   a, 
>   SUM(b) OVER (PARTITION BY c ORDER BY procTime() RANGE BETWEEN UNBOUNDED PRECEDING AND
CURRENT ROW) AS sumB,
>   MIN(b) OVER (PARTITION BY c ORDER BY procTime() RANGE BETWEEN UNBOUNDED PRECEDING AND
CURRENT ROW) AS minB
> FROM myStream
> {code}
> The following restrictions should initially apply:
> - All OVER clauses in the same SELECT clause must be exactly the same.
> - The PARTITION BY clause is optional (no partitioning results in single threaded execution).
> - The ORDER BY clause may only have procTime() as parameter. procTime() is a parameterless
scalar function that just indicates processing time mode.
> - bounded PRECEDING is not supported (see FLINK-5654)
> - FOLLOWING is not supported.
> The restrictions will be resolved in follow up issues. If we find that some of the restrictions
are trivial to address, we can add the functionality in this issue as well.
> This issue includes:
> - Design of the DataStream operator to compute OVER ROW aggregates
> - Translation from Calcite's RelNode representation (LogicalProject with RexOver expression).



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Mime
View raw message