[
https://issues.apache.org/jira/browse/BEAM-14244?focusedWorklogId=753404&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-753404
]
ASF GitHub Bot logged work on BEAM-14244:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 06/Apr/22 13:48
Start Date: 06/Apr/22 13:48
Worklog Time Spent: 10m
Work Description: steveniemitz commented on PR #17262:
URL: https://github.com/apache/beam/pull/17262#issuecomment-1090295080
> LGTM. Do you plan to do the follow-up for watermark hold in Flink? Would
you do a `@ValidatesRunner` test for that?
If I have some time I can take a swing at it, but I don't think it'd be for
a little while. I also don't use Flink so verifying that it actually works
correctly will be more tricky. :)
On the surface the fix looks pretty easy, just adding a watermark hold in
DoFnOperator for processing time timers as it does for event time ones (and
clearing them as well).
Issue Time Tracking
-------------------
Worklog Id: (was: 753404)
Time Spent: 4h (was: 3h 50m)
> Processing time timers should use outputTimestamp rather than input watermark
> for their timestamp
> -------------------------------------------------------------------------------------------------
>
> Key: BEAM-14244
> URL: https://issues.apache.org/jira/browse/BEAM-14244
> Project: Beam
> Issue Type: Bug
> Components: runner-core, sdk-java-core
> Reporter: Steve Niemitz
> Assignee: Steve Niemitz
> Priority: P1
> Time Spent: 4h
> Remaining Estimate: 0h
>
> Currently processing time timers ignore the outputTimestamp and instead use
> the input watermark at the time they fire. This is wrong because the input
> watermark can have advanced arbitrarily far past the actual output timestamp
> when it fires.
> The correct behavior should be to instead use the outputTimestamp the timer
> was configured to fire with.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
