[
https://issues.apache.org/jira/browse/FLINK-4082?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15341508#comment-15341508
]
ASF GitHub Bot commented on FLINK-4082:
---------------------------------------
Github user uce commented on the issue:
https://github.com/apache/flink/pull/2113
Changes look good, but did you skip the `CombiningUnilateralSortMerger` on
purpose?
It's hard to tell, because the `handleLargeRecords` argument was existing
before your changes, but it was not configurable. The
`CombiningUnilateralSortMerger` always sets `handleLargeRecords` to `true`.
> Add Setting for LargeRecordHandler
> ----------------------------------
>
> Key: FLINK-4082
> URL: https://issues.apache.org/jira/browse/FLINK-4082
> Project: Flink
> Issue Type: Improvement
> Reporter: Aljoscha Krettek
> Assignee: Aljoscha Krettek
>
> Now, this is always enabled but there are known problems when users specify a
> custom {{TypeInformation}}. We should introduce a setting for this.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
