[
https://issues.apache.org/jira/browse/FLINK-38738?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Timo Walther reopened FLINK-38738:
----------------------------------
It turned out that updating connectors is still hard if they don't understand
in which version they operate.
> Incorrect UID for DataStreamScanProvider and TransformationScanProvider
> -----------------------------------------------------------------------
>
> Key: FLINK-38738
> URL: https://issues.apache.org/jira/browse/FLINK-38738
> Project: Flink
> Issue Type: Bug
> Components: Table SQL / Runtime
> Reporter: Arvid Heise
> Assignee: Timo Walther
> Priority: Critical
> Labels: pull-request-available
> Fix For: 2.3.0
>
>
> The planner guarantees that all operators receive a stable UID. It does this,
> by deriving the UIDs from the CompiledPlan. Thus, CompiledPlan and
> savepoint/checkpoint are closely linked.
> The planner does not control the operators and transformations in sources and
> sinks. From a planner’s perspective, sources and sinks are black boxes.
> Implementers must ensure that UIDs are assigned to all operators. The planner
> provides utilities for generating UIDs:
> https://github.com/apache/flink/blob/master/flink-table/flink-table-common/src/main/java/org/apache/flink/table/connector/ProviderContext.java
> Even though a source generates and assigns UIDs for all its operators. The
> planner overwrites the UID of the last transformation with
> <ExecNoceId>_source. For example, given a source that adds two
> transformations: an operator {{4_kafka}} followed by an operator
> {{4_splitting}}, effectively leads to an operator {{4_kafka}} followed by an
> operator {{4_source}}. Or a source that adds one transformation, an operator
> {{4_kafka}}, effectively leads to an operator {{4_source}}.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
