Hey everyone!
I’m dealing with a tricky Debezium PostgreSQL connector issue and could use some advice.
The ProblemMy PostgreSQL DB was converted from Oracle using AWS Schema Conversion Tool, and it has Oracle compatibility extensions installed. This created 40K+ custom types (yes, really).
When I try to run Debezium, the connector gets stuck during startup because it’s processing all of these types. The logs keep filling up with messages like:
WARN Type [oid:316992, name:some_oracle_type] is already mapped
WARN Type [oid:337428, name:another_type] is already mapped
It’s been churning on this for hours.
My SetupPostgreSQL 13 with Oracle compatibility extensions
Kafka Connect in Docker
Only want to capture CDC from one schema and one table
Current config (simplified):
include.unknown.datatypes=false (but then connector fails)
errors.tolerance=all, errors.log.enable=true
Filters to only include the schema + table I need
Excluding unknown data types → connector won’t start
Adding error tolerance configs → no effect
Schema/table filters → still stuck on type discovery
Has anyone here dealt with Debezium + Oracle compatibility extensions before?
Is there a way to skip type discovery for schemas/tables I don’t care about?
Would I be better off creating a clean PostgreSQL DB without Oracle extensions and just migrating my target schema?
Are there specific Debezium configs for handling this scenario?
The connector technically starts (tasks show up in logs), but it’s unusable because it’s processing thousands of types I don’t need.
Any tips, workarounds, or war stories would be greatly appreciated! 🙏 help a brother out bean stuck on it for 2 weeks
--
You received this message because you are subscribed to the Google Groups "debezium" group.
To unsubscribe from this group and stop receiving emails from it, send an email to debezium+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/debezium/c5bfd4ed-87d4-43ee-8008-0b183f693947n%40googlegroups.com.