自增长在并行插入中如何工作？

在并行批量插入中使用自增长可能会导致自增长列之间的行号冲突。为避免这种情况，可以使用分段技术来使每个线程插入唯一的自增长值。以下是一个示例代码：

import threading
import pymysql

class BulkInsertThread(threading.Thread):
  def __init__(self, start_id, end_id):
    threading.Thread.__init__(self)
    self.start_id = start_id
    self.end_id = end_id

  def run(self):
    connection = pymysql.connect(user='user', password='password', host='localhost', database='database')
    cursor = connection.cursor()
    for i in range(self.start_id, self.end_id):
      cursor.execute("INSERT INTO table_name (column_name) VALUES (%s)", (i,))
    connection.commit()
    cursor.close()
    connection.close()

def bulk_insert():
  NUM_THREADS = 4
  ROWS_PER_THREAD = 1000
  for i in range(NUM_THREADS):
    start_id = i * ROWS_PER_THREAD + 1
    end_id = (i + 1) * ROWS_PER_THREAD + 1
    thread = BulkInsertThread(start_id, end_id)
    thread.start()

bulk_insert()

该示例使用了4个线程并行执行插入操作，每个线程插入1000行数据。每个线程插入的行号都是唯一的，避免了自增长冲突的问题。