[
https://issues.apache.org/jira/browse/FLINK-6069?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15944856#comment-15944856
]
Kostas Kloudas commented on FLINK-6069:
---------------------------------------
Hi [~patricklx] ,
Currently the CEP library had a somehow fuzzy way to handle late events.
Essentially:
1) it was accepting all events (late and early)
2) whenever a watermark was arriving it was sorting them based on event time
3) it was feeding them into the NFA.
As you see this is not respecting event time.
I will open a PR soon (already working on it) where the underlying assumption
is that the watermark is correct. This means that late events are discarded.
Late events are the ones whose timestamp is smaller than the last received
watermark.
Dropped events will be sent to a side output, as done in the case of the
{{WindowOperator}}.
> Add documentation CEP - Watermarks
> ----------------------------------
>
> Key: FLINK-6069
> URL: https://issues.apache.org/jira/browse/FLINK-6069
> Project: Flink
> Issue Type: Improvement
> Components: CEP
> Affects Versions: 1.2.0, 1.3.0
> Reporter: Patrick Pircher
> Assignee: Kostas Kloudas
> Priority: Minor
>
> I was trying to understand how CEP works with EventTime/Watermarks and
> specifically how it handles events that come out of order.
> So the only thing I found was an email archive and some pointers in the
> Source Code.
> http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/CEP-and-slightly-out-of-order-elements-td9439.html
> https://github.com/apache/flink/blob/release-1.2.0/flink-libraries/flink-cep/src/main/java/org/apache/flink/cep/operator/AbstractCEPPatternOperator.java#L71
> https://github.com/apache/flink/blob/release-1.2.0/flink-libraries/flink-cep/src/main/java/org/apache/flink/cep/operator/AbstractCEPPatternOperator.java#L96
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
