The new connector should handle that situation just fine.
If you want to use filesystem storage, it’s possible to configure each connector instance to use a different offset storage file.
If you want to use the Kafaka storage, the bucket name is part of the namespace for the offset storage. As long as the connectors all read from different buckets, their offset storage will be separate.
If you wanted multiple connectors to read from the same bucket, there’s an additional config property
compat.connector_name_in_offsets that must be set to
true. But it sounds like that should be not be necessary in your case, since each connector instance reads from a different bucket.
Again, I’m not completely familiar with the features and limitations of the old connector, so it’s possible I’ve misunderstood your situation. If that’s the case, let me know