[
https://issues.apache.org/jira/browse/FLINK-1967?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14611693#comment-14611693
]
ASF GitHub Bot commented on FLINK-1967:
---------------------------------------
Github user StephanEwen commented on the pull request:
https://github.com/apache/flink/pull/879#issuecomment-117974230
Concerning watermarks and performance: Have we considered sending
watermarks as records, rather than events?
- Events in data channels are inevitably more heavyweight than records.
They have less efficient serialization and flush buffers.
- In cases where we can guarantee that records have ascending time stamps
in a stream (sub) partition, every record is a watermark by itself, and there
is no need to have extra watermarks.
The downside of having records rather than events is that we sometimes
really want to flush the buffers with a watermark. On the other hand, buffers
are flushed periodically anyways, so no big deal, probably.
If our stream records are always 2-tuples (timestamp, record) , then we
could do something like shipping (timestamp, null) records as watermarks.
> Introduce (Event)time in Streaming
> ----------------------------------
>
> Key: FLINK-1967
> URL: https://issues.apache.org/jira/browse/FLINK-1967
> Project: Flink
> Issue Type: Improvement
> Reporter: Aljoscha Krettek
> Assignee: Aljoscha Krettek
>
> This requires introducing a timestamp in streaming record and a change in the
> sources to add timestamps to records. This will also introduce punctuations
> (or low watermarks) to allow windows to work correctly on unordered,
> timestamped input data. In the process of this, the windowing subsystem also
> needs to be adapted to use the punctuations. Furthermore, all operators need
> to be made aware of punctuations and correctly forward them. Then, a new
> operator must be introduced to to allow modification of timestamps.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
