Problems with batch inserts

[Guitlle]

Hi. I am trying to replicate a mysql table using the batch inserts. Here is the code I've written for this:

mysql_cursor = self.mysql.cursor() mysql_cursor.execute("SELECT * FROM" + self.table + "` WHERE " + where)
column_names = [i[0] for i in mysql_cursor.description]

    batch = VerticaBatch(
        table_name= self.vertica_schema + '.' + self.table,
        truncate_table=True,
        column_list=column_names,
        copy_options={
            'DELIMITER': ',',
        },
        connection = self.vertica,
        analyze_constraints= False,
        multi_batch= True
    )

    count = 0
    while True:
        if count % batch_size == 0:
            if count > 0:
                print(count, mysql_cursor.rowcount)
                error_bool, error_file_obj = batch.get_errors()
                if error_bool:
                    # TODO: Do something with this error
                    error = error_file_obj.read()
                batch.commit()
        count += 1
        msrow = mysql_cursor.fetchone()
        if msrow is None:
            break;
        batch.insert_list(list(msrow))

    error_bool, error_file_obj = batch.get_errors()
    if error_bool:
       # TODO: Do something with this error
       error = error_file_obj.read()
    batch.commit()
    batch.close_batch()

`

Some times, it logs the warning "Error shutting down task thread!".

When I test this code on an iPython console for the first time, it works well. If I try to make a new class and run this a second time, it hangs after the first batch.commit() call.

Is it possible that threads and pyodbc are buggy? I will try to rewrite pyvertica.batch without threads to see if things get better.