[
https://issues.apache.org/jira/browse/FLINK-16693?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17068770#comment-17068770
]
Jark Wu commented on FLINK-16693:
---------------------------------
Yes. You are right [~Paul Lin]. I didn't realize the side effect to the old
planner. If a connector always use LocalDateTime, it can work in blink planner,
but will not work in old planner, because old planner doesn't support
LocalDateTime...
The fixing will be rather difficult, because it's just about the validation in
{{TableSinkUtils}}, but the whole planner should support LocalDateTime,
including code generation...
> Legacy planner incompatible with Timestamp backed by LocalDateTime
> ------------------------------------------------------------------
>
> Key: FLINK-16693
> URL: https://issues.apache.org/jira/browse/FLINK-16693
> Project: Flink
> Issue Type: Bug
> Components: Table SQL / Legacy Planner
> Affects Versions: 1.10.0
> Reporter: Paul Lin
> Priority: Major
>
> Recently I upgraded a simple application that inserts static data into a
> table from 1.9.0 to 1.10.0, and
> encountered a timestamp type incompatibility problem during the table sink
> validation.
> The SQL is like:
> ```
> insert into kafka.test.tbl_a # schema: (user_name STRING, user_id INT,
> login_time TIMESTAMP)
> select ("ann", 1000, TIMESTAMP "2019-12-30 00:00:00")
> ```
> And the error thrown:
> ```
> Field types of query result and registered TableSink `kafka`.`test`.`tbl_a`
> do not match.
> Query result schema: [EXPR$0: String, EXPR$1: Integer, EXPR$2:
> Timestamp]
> TableSink schema: [user_name: String, user_id: Integer, login_time:
> LocalDateTime]
> ```
> After some digging, I found the root cause might be that since FLINK-14645
> timestamp fields defined via TableFactory had been bridged to LocalDateTime,
> but timestamp functions are still backed by java.sql.Timestamp.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
