[jira] [Assigned] (FLINK-34783) [Bug] TiDB Flink CDC not support exactly once

["Leonard Xu (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/FLINK-34783?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Leonard Xu reassigned FLINK-34783:
----------------------------------

    Assignee: ouyangwulin

> [Bug] TiDB Flink CDC not support exactly once
> ---------------------------------------------
>
>                 Key: FLINK-34783
>                 URL: https://issues.apache.org/jira/browse/FLINK-34783
>             Project: Flink
>          Issue Type: Bug
>          Components: Flink CDC
>            Reporter: Flink CDC Issue Import
>            Assignee: ouyangwulin
>            Priority: Major
>              Labels: github-import
>
> ### Search before asking
> - [X] I searched in the [issues|https://github.com/ververica/flink-cdc-connectors/issues) and found nothing similar.
> ### Flink version
> 1.13.6
> ### Flink CDC version
> I had try three version:
> 2.3.0
> 2.4-SNAPSHOT
> 2.2.1
> ### Database and its version
> I had try tow version:
> tidb v5.4.3
> tidb v6.1.3
> ### Minimal reproduce step
> (1)Start a Flink CDC program locally, read a table in tidb, and enable checkpoints in the Flink program to remain in a directory (I tested the Flink program locally, so the checkpoints save location uses the local directory).
> (2)Stop this Flink CDC program to simulate an exception scenario, and then insert, delete, and update this table in TiDB.
> (3]Restoring the Flink cdc program from checkpoints will trace the operation of the Tidb table after the program experienced an exception. Insertion and update operations on the Tidb table can be restored to the Flink program for processing, but the deletion operation on the Tidb table has been lost.
> ### What did you expect to see?
> Restoring the Flink cdc program from checkpoints will trace the operations of the Tidb table after the program experienced an exception. Insertion, update, and deletion operations on the Tidb table can be restored to the Flink program for processing, which means that the deletion operation will not be lost.
> ### What did you see instead?
> Restoring the Flink cdc program from checkpoints will trace the operation of the Tidb table after the program experienced an exception. Insertion and update operations on the Tidb table can be restored to the Flink program for processing, but the deletion operation on the Tidb table has been lost.
> ### Anything else?
> Does Flink cdc tidb really implement end-to-end exactly once?
> ### Are you willing to submit a PR?
> - [x] I'm willing to submit a PR!
> ---------------- Imported from GitHub ----------------
> Url: https://github.com/apache/flink-cdc/issues/2034
> Created by: [ArainChen|https://github.com/ArainChen]
> Labels: bug, 
> Created at: Wed Mar 22 15:59:22 CST 2023
> State: open



--
This message was sent by Atlassian Jira
(v8.20.10#820010)