[ 
https://issues.apache.org/jira/browse/FLINK-8144?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16267762#comment-16267762
 ] 

ASF GitHub Bot commented on FLINK-8144:
---------------------------------------

Github user fhueske commented on the issue:

    https://github.com/apache/flink/pull/5063
  
    A punctuated watermark assigner can emit `null` instead of a watermark. So 
using a punctuated watermark assigner does not mean that for each event a 
watermark must be emitted. The motivation for punctuated watermarks is to react 
to specific events that carry watermark information and not to all events (see 
[JavaDocs](https://github.com/apache/flink/blob/master/flink-streaming-java/src/main/java/org/apache/flink/streaming/api/functions/AssignerWithPunctuatedWatermarks.java)).
 Of course there's nothing that prevents users from emitting one watermark per 
event but as I said this was not the intended use case.
    
    A `lastProcessedWatermark` state variable would improve the situation a bit 
by acting as an early out for the `onTimer()` method. However, the system would 
still need to checkpoint and handle significantly more timers than with the 
current implementation.


> Optimize the timer logic in RowTimeUnboundedOver
> ------------------------------------------------
>
>                 Key: FLINK-8144
>                 URL: https://issues.apache.org/jira/browse/FLINK-8144
>             Project: Flink
>          Issue Type: Bug
>          Components: Table API & SQL
>            Reporter: Dian Fu
>            Assignee: Dian Fu
>             Fix For: 1.5.0
>
>
> Currently the logic of {{RowTimeUnboundedOver}} is as follows:
> 1) When element comes, buffer it in MapState and and register a timer at 
> {{current watermark + 1}}
> 2) When event timer triggered, scan the MapState and find the elements below 
> the current watermark and process it. If there are remaining elements to 
> process, register a new timer at {{current watermark + 1}}.
> Let's assume that watermark comes about 5 seconds later than the event on 
> average, then we will scan about 5000 times the MapState before actually 
> processing the events.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to