[
https://issues.apache.org/jira/browse/FLINK-3669?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15223402#comment-15223402
]
Konstantin Knauf edited comment on FLINK-3669 at 4/3/16 5:32 PM:
-----------------------------------------------------------------
I tried to implement the 1.. The hard part, I think, is to store the
{{ScheduledFuture}}s to in the {{WindowOperator}}. We could use a
{{Map<Long,ScheduledFuture>}}, but this only works if we also implement timer
coalescing. Besides that I am unsure about how to de/serialize this map. Could
you give me pointer?
Edit. Ok, serializing the future does not make sense, so it could not be
checkpointed, which is not necessarily a problem, I think. Additionally, this
should be a {{Map<Timer, ScheduledFuture>}} instead.
was (Author: knaufk):
I tried to implement the 1.. The hard part, I think, is to store the
{{ScheduledFuture}}s to in the {{WindowOperator}}. We could use a
{{Map<Long,ScheduledFuture>}}, but this only works if we also implement timer
coalescing. Besides that I am unsure about how to de/serialize this map. Could
you give me pointer?
> WindowOperator registers a lot of timers at StreamTask
> ------------------------------------------------------
>
> Key: FLINK-3669
> URL: https://issues.apache.org/jira/browse/FLINK-3669
> Project: Flink
> Issue Type: Bug
> Components: Streaming
> Affects Versions: 1.0.1
> Reporter: Aljoscha Krettek
> Priority: Blocker
>
> Right now, the WindowOperator registers a timer at the StreamTask for every
> processing-time timer that a Trigger registers. We should combine several
> registered trigger timers to only register one low-level timer (timer
> coalescing).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
