I still getting 'NoiSplitter' object has no attribute 'source', even on version 1.8.2.
In order to get it running, I had to do a makeshift change in the code of splitter.py:

class NoSplitter(BaseSplitter):
    """No split bounded source so not work parallel."""

    def estimate_size(self):
        # return self.source.client.rough_counts_estimator(self.source.query)
        return 0x100000 # may be any number

I get this error when using apache-beam 2.17.0 and 2.18.0.

None values are converted into the string 'None' instead of NULL.
The bug can be fixed using this patch:

--- a/beam_mysql/connector/io.py
+++ b/beam_mysql/connector/io.py
@@ -115,7 +115,8 @@ class _WriteToMySQLFn(beam.DoFn):
             values.append(value)

         column_str = ", ".join(columns)
-        value_str = ", ".join([f"{value}" if isinstance(value, (int, float)) else f"'{value}'" for value in values])
+        value_str = ", ".join([f"{'NULL' if value is None else value}" if isinstance(value, (type(None), int, float))
+                               else f"'{value}'" for value in values])
         query = f"INSERT INTO {self._config['database']}.{self._table}({column_str}) VALUES({value_str});"

         self._queries.append(query)
--

Hello! I just ran the example code in beam-mysql-connector/examples/write_records_pipeline.py on my computer in order to see if I can write into MariaDB table. It returned an error: _mysql_connector.MySQLInterfaceError: Commands out of sync; you can't run this command now [while running 'WriteToMySQL/ParDo(_WriteToMySQLFn)'].

However, if I run the same code with only one row in beam.Create([...]), like beam.Create([{"name": "test data3"}]), it will work just fine.

Is it possible to avoid somehow this type of error using beam-mysql-connector and load multiple rows?

Hi,
I am using the connector to write to MySql using batch size as 1000,but I get error suggesting to use bulk Insert with cmd_insert option.
Could you please share few examples with bulk Insert?The example provided uses insert option with batch size as 0.
Thanks!