我使用BigQuery作为我的数据库与Scrapy蜘蛛。下面是将数据存储到DB中的两个管道。一个使用Insert，另一个使用Update方法。Update方法比insert慢70 times（每分钟仅更新20条记录）。更新需要3.560秒，而插入只需0.05秒。我错在哪里，如何加快更新方法？
P.S.当前表的大小大约是20k条记录，潜在的大小可以高达500000条记录。需要每天更新记录。

更新方式

# Define the update query
query = f"""
        UPDATE `{self.dataset_id}.{self.table_id}`
        SET `Sold Status` = '{data['Sold Status']}',
            `Amount of Views` = '{data['Amount of Views']}',
            `Amount of Likes` = '{data['Amount of Likes']}',
            `Sold Date & Time` = '{data['Sold Date & Time']}'
        WHERE `Item number` = '{data['Item number']}'
    """
start_time = time.time()
# Run the update query
job = self.client.query(query)

# Wait for the job to complete
job.result()

# Check if the query was successful
if job.state == 'DONE':
    print('Update query executed successfully.')
else:
    print('Update query failed.')
end_time = time.time()
execution_time = end_time - start_time
logging.info(execution_time)

return item

插入方式

start_time = time.time()
data = item
slug = data['slug']
if slug in self.ids_seen:
    raise DropItem("Duplicate item found: {}".format(slug))
else:
    data.pop('slug', None)
    self.ids_seen.add(slug)
    table_ref = self.client.dataset(self.dataset_id).table(self.table_id)

    # Define the rows to be inserted
    rows = [
        data
    ]

    # Insert rows into the table
    errors = self.client.insert_rows_json(table_ref, rows)

    if errors == []:
        print("Rows inserted successfully.")
    else:
        print("Encountered errors while inserting rows:", errors) 
    end_time = time.time()
    execution_time = end_time - start_time
    logging.info(execution_time)
    
    return item