SEZ9 commented on issue #12267:
URL: https://github.com/apache/seatunnel/issues/12267#issuecomment-5707498099

   Thanks @1362227089, this is the detail we needed.
   
   The error text answers the database/schema question: the connector reports 
the table as `highgo.testdb.test_a`, which matches the database, schema and 
table you configured, so the table was located correctly and this does not look 
like a database-name/schema-name confusion. The check is specifically asking 
for a full replica identity, and the table currently has `relreplident = 'd'`. 
A primary key with the default identity is exactly what this message rejects, 
so the failing condition is the replica identity setting, not the primary key.
   
   The `TableDiscoveryUtils` warnings for `template1`, `template0` and `cflag` 
only show those databases being skipped during discovery; the failing check is 
the replica-identity one above.
   
   Remaining asks so we can close this out:
   
   1. Run the statement from the message, `ALTER TABLE highgo.testdb.test_a 
REPLICA IDENTITY FULL;`, confirm the new `relreplident` value, and re-run the 
same job with the standard PostgreSQL driver and the MD5 CDC user, keeping 
ports and connector code unchanged.
   2. The pasted log ends mid-line ("Connection closed by") and does not 
include the exception stack trace itself. Please attach the full sanitized 
stack trace of the failing run, plus the startup-mode value used.
   3. If the job gets past initialization, report the result of one 
snapshot-only run and one logical-replication run against `testdb.test_a` 
(slot/publication creation and whether WAL changes arrive).
   
   If the replica-identity change lets the job start, we are back to a normal 
Postgres-CDC configuration on that server/auth combination. If it fails at a 
later step, that is where a HighGo-specific logical-replication difference 
would start to show, and we can evaluate it from there.
   
   <!-- streview-comment:1108 -->


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to