The idea is to provide sth similar to a byte[] (byte array) data type.
Example of SQL Server:
- binary [ ( n ) ] Fixed-length binary data with a length of n bytes, where n is a value from 1 through 8,000. The storage size is n bytes.
- varbinary [ ( n | max) ] Variable-length binary data. n can be a value from 1 through 8,000. max indicates that the maximum storage size is 2^31-1 bytes (2 GB). The storage size is the actual length of the data entered + 2 bytes. The data that is entered can be 0 bytes in length. The ANSI SQL synonym for varbinary is binary varying.
We already have #1282
For reference, https://github.com/confluentinc/ksql/issues/1282 is "Add support for the "bytes" type in avro".
Should we close one of them as a duplicate? This one seems to have more context and is more general.
Yes, I closed https://github.com/confluentinc/ksql/issues/1282 as a duplicate of this one.
This would be help for me when creating a stream over an Oracle Golden Gate topic containing image data - I don't need to process it, just pass it as is to the consumer
+1 for this. Use case: we would like to have the key as a binary, and then have the JDBC connector sink the topic into MySQL, where the PK is a binary.
+1 I've got a Debezium connector reading stuff in from Oracle. Some data comes through onto the topic as a byte array. I'd like to have a custom UDF to turn this into something more useful. However, I can't see the data and I believe it is because KSQL doesn't yet play with byte arrays.
@apurvam @miguno Do we have any update for using binaries such as JSONB in KSQL or KSQLDB ?
+1 I've got Oracle data coming in and want to easily separate out the BLOB data from the rest of the record.
Most helpful comment
This would be help for me when creating a stream over an Oracle Golden Gate topic containing image data - I don't need to process it, just pass it as is to the consumer