[
https://issues.apache.org/jira/browse/FLINK-39252?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18066236#comment-18066236
]
Daishuyuan commented on FLINK-39252:
------------------------------------
Thanks for the review.
This branch now includes:
- the initial SQL Server pipeline connector implementation
- SQL Server `snapshot` / `timestamp` startup mode support
- follow-up fixes for:
- expired CDC window handling
- rejecting `timestamp` mode on the legacy non-incremental path
- avoiding duplicate `CreateTableEvent` emission after restore/rescale
- fixing `listSchemas(null)` fallback behavior
- aligning the `tables` option description with the current implementation
This passed locally.
Could you please take another look at the latest branch state?
> Flink cdc pipeline support sqlserver source pipeline connector
> --------------------------------------------------------------
>
> Key: FLINK-39252
> URL: https://issues.apache.org/jira/browse/FLINK-39252
> Project: Flink
> Issue Type: Improvement
> Components: Flink CDC
> Environment: * Flink version: 1.20.3
> * Debezium version: 1.9.8.Final
> Reporter: Daishuyuan
> Priority: Major
> Original Estimate: 336h
> Remaining Estimate: 336h
>
> * Motivation
> Flink CDC already provides a SQL Server source connector, but the pipeline
> connector side is still missing. This makes SQL Server inconsistent with the
> other databases already supported by the pipeline framework, and prevents
> users from using SQL Server CDC in the unified pipeline mode.
> * What is proposed
> This change adds a new SQL Server pipeline connector module and integrates it
> into the pipeline connectors build.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
