[
https://issues.apache.org/jira/browse/FLINK-7193?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Flink Jira Bot updated FLINK-7193:
----------------------------------
Labels: auto-deprioritized-major auto-deprioritized-minor (was:
auto-deprioritized-major stale-minor)
Priority: Not a Priority (was: Minor)
This issue was labeled "stale-minor" 7 days ago and has not received any
updates so it is being deprioritized. If this ticket is actually Minor, please
raise the priority and ask a committer to assign you the issue or revive the
public discussion.
> Update all OVER window ProcessFunction implementations to use SortedQueueState
> ------------------------------------------------------------------------------
>
> Key: FLINK-7193
> URL: https://issues.apache.org/jira/browse/FLINK-7193
> Project: Flink
> Issue Type: Improvement
> Components: Table SQL / API
> Affects Versions: 1.3.1, 1.4.0
> Reporter: Fabian Hueske
> Priority: Not a Priority
> Labels: auto-deprioritized-major, auto-deprioritized-minor
>
> Most OVER windows are implemented as {{ProcessFunction}} with {{MapState}}
> which is keyed on a timestamp. The algorithms are implemented in a way that
> they can leverage a sorted traversal over the timestamp keys but fall back to
> sorting if the traversed keys are not sorted.
> The {{RocksDBStateBackend}} provides sorted iterators.
> FLINK-6219 proposes a {{SortedQueueState}} that guarantees that values can be
> read from a sorted list. Once FLINK-6219 is resolved an a
> {{SortedQueueState}} is available, the OVER window implementations should be
> updated to use the new state type instead of {{MapState}}.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
