thswlsqls opened a new issue, #8604: URL: https://github.com/apache/paimon/issues/8604
**Search before asking** - [x] I searched in the [issues](https://github.com/apache/paimon/issues) and found nothing similar. **Paimon version** master @ ec1d8ee8a / 2.0-SNAPSHOT **Compute Engine** Flink (CDC ingestion — PostgreSQL sync) **Minimal reproduce step** Run `postgres_sync_table`/`postgres_sync_database` on a table with a `time(n)` column where `n <= 3`. Debezium's default adaptive mode encodes it as the `io.debezium.time.Time` logical type (int32, millis-of-day). `PostgresRecordParser.extractFieldType()` (paimon-flink-cdc `.../postgres/PostgresRecordParser.java` line 159-163) handles only `io.debezium.time.Date` in the int32 branch and falls through to `INT()`. **What doesn't meet your expectations?** Expected: the column maps to `TIME(3)`, matching the JDBC bootstrap path `PostgresTypeUtils.toDataType()` which maps `time` to `TIME(scale)`. Actual: it maps to `INT`, so the runtime and bootstrap schema paths disagree and the sync crashes on the first record with "Cannot convert field from TIME(3) to INT", looping on restart. **Anything else?** Incomplete-fix counterpart of merged PR #8222, which fixed the symmetric int64 `Timestamp` case. The int64 branch already maps `Timestamp`/`MicroTimestamp`/`MicroTime`; `time(4..6)` (`MicroTime`) is already handled. **Are you willing to submit a PR?** - [x] I'm willing to submit a PR! -- 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]
