[ 
https://issues.apache.org/jira/browse/CASSANALYTICS-182?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Jyothsna Konisa reassigned CASSANALYTICS-182:
---------------------------------------------

    Assignee: Jyothsna Konisa

> CDC batch-write mixing a CDC-enabled and CDC-disabled table drops the CDC 
> table's mutation
> ------------------------------------------------------------------------------------------
>
>                 Key: CASSANALYTICS-182
>                 URL: https://issues.apache.org/jira/browse/CASSANALYTICS-182
>             Project: Apache Cassandra Analytics
>          Issue Type: Bug
>          Components: CDC
>            Reporter: Jyothsna Konisa
>            Assignee: Jyothsna Konisa
>            Priority: Normal
>
> Problem: A BEGIN BATCH statement writing to both a CDC-enabled and a 
> CDC-disabled table under the same partition key produces one combined 
> commit-log entry. Because the CDC bridge only tracked CDC-enabled tables, it 
> couldn't recognize the
>   non-CDC table in that entry, threw an error, and dropped the entire batch — 
> including the CDC-enabled table's own update. Net effect: silent CDC data 
> loss.
> Fix: The CDC bridge can now register any table, not just CDC-enabled ones, 
> each correctly tagged with whether CDC is actually on for it. It's the caller 
> that decides which non-CDC tables actually need registering — not every table 
> in the cluster, only those that could realistically end up batched with a 
> CDC-enabled table. Publishing/filtering logic still only acts on CDC-enabled 
> tables, so batches deserialize correctly and no CDC events get silently 
> dropped, without registering more schema than necessary.



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

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to