Python日志处理：串行与并行执行优化及Logstash JSON格式配置

唐可

发布时间：2026-10-13 | 4人浏览过

原创

本文介绍Python中串行与并行执行的差异，重点讲解如何通过Python生成JSON格式日志以优化Logstash性能。通过对比直接发送与文件Tail两种方案，以及python-logstash模块的AMQP支持，帮助开发者解决日志字段解析问题，提升ELK栈处理效率。

代码示例如下：

Python串行与并行执行示例


#coding=utf-8
 
import threading
 
import time
 
import cx_Oracle
 
from pprint import pprint
 
import csv
 
print time.asctime()
 
table_name = "dbtest.csv"
 
f = open(table_name + ".csv", "w")
 
conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
cursor = conn.cursor()
 
def query01():
 
  tname = threading.current_thread()
 
  aa=10
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
def query02():
 
  tname = threading.current_thread()
 
  aa=20
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
def query03():
 
  tname = threading.current_thread()
 
  aa=30
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
def query04():
 
  tname = threading.current_thread()
 
  aa=40
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
def query05():
 
  tname = threading.current_thread()
 
  aa=50
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
def query06():
 
  tname = threading.current_thread()
 
  aa=60
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
def query07():
 
  tname = threading.current_thread()
 
  aa=70
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  return row
 
  cursor.close()
 
def query08():
 
  tname = threading.current_thread()
 
  aa=80
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
def query09():
 
  tname = threading.current_thread()
 
  aa=90
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
def query10():
 
  tname = threading.current_thread()
 
  aa=100
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close() 
 
 
threads = []
 
t1 = threading.Thread(target=query01, name='query01')
 
threads.append(t1)
 
t2 = threading.Thread(target=query02, name='query02')
 
threads.append(t2)
 
t2 = threading.Thread(target=query03, name='query03')
 
threads.append(t2)
 
t2 = threading.Thread(target=query04, name='query04')
 
threads.append(t2)
 
t2 = threading.Thread(target=query05, name='query05')
 
threads.append(t2)
 
t2 = threading.Thread(target=query06, name='query06')
 
threads.append(t2)
 
t2 = threading.Thread(target=query07, name='query07')
 
threads.append(t2)
 
t2 = threading.Thread(target=query08, name='query08')
 
threads.append(t2)
 
t2 = threading.Thread(target=query09, name='query09')
 
threads.append(t2)
 
t2 = threading.Thread(target=query10, name='query10')
 
threads.append(t2) 
 
if __name__ == '__main__':
 
  for t in threads:
 
    # t.setDaemon(True)
 
    t.start()
 
    # t.run()
 
    # t.start()
 
    # print '3333333'
 
    print threading.current_thread()
 
    # print t.is_alive()
 
    # print '3333333'
 
    t.join()
 
  print "all over "
 
  print time.asctime() 
 
 
C:\Python27\python.exe C:/Users/tlcb/PycharmProjects/untitled/a2.py
 
Wed Mar 28 11:08:19 2018
 
<_MainThread(MainThread, started 18744)>
 
[(10, '10boobook10', '10aaaaaaaaaaaa10', '10bbbbbbbbbbbbbbbbb10'), (10, '10sssssssss10', 'tlcb', 'tlcb'), (10, '10boobook10', '10aaaaaaaaaaaa10', '10bbbbbbbbbbbbbbbbb10')]
 
<_MainThread(MainThread, started 18744)>
 
[(20, '20boobook20', '20aaaaaaaaaaaa20', '20bbbbbbbbbbbbbbbbb20'), (20, '20boobook20', '20aaaaaaaaaaaa20', '20bbbbbbbbbbbbbbbbb20'), (20, '20boobook20', '20aaaaaaaaaaaa20', '20bbbbbbbbbbbbbbbbb20')]
 
<_MainThread(MainThread, started 18744)>
 
[(30, '30boobook30', '30aaaaaaaaaaaa30', '30bbbbbbbbbbbbbbbbb30'), (30, '30boobook30', '30aaaaaaaaaaaa30', '30bbbbbbbbbbbbbbbbb30'), (30, '30boobook30', '30aaaaaaaaaaaa30', '30bbbbbbbbbbbbbbbbb30')]
 
<_MainThread(MainThread, started 18744)>
 
[(40, '40boobook40', '40aaaaaaaaaaaa40', '40bbbbbbbbbbbbbbbbb40'), (40, '40boobook40', '40aaaaaaaaaaaa40', '40bbbbbbbbbbbbbbbbb40'), (40, '40boobook40', '40aaaaaaaaaaaa40', '40bbbbbbbbbbbbbbbbb40')]
 
<_MainThread(MainThread, started 18744)>
 
[(50, '50boobook50', '50aaaaaaaaaaaa50', '50bbbbbbbbbbbbbbbbb50'), (50, '50boobook50', '50aaaaaaaaaaaa50', '50bbbbbbbbbbbbbbbbb50'), (50, '50boobook50', '50aaaaaaaaaaaa50', '50bbbbbbbbbbbbbbbbb50')]
 
<_MainThread(MainThread, started 18744)>
 
[(60, '60boobook60', '60aaaaaaaaaaaa60', '60bbbbbbbbbbbbbbbbb60'), (60, '60boobook60', '60aaaaaaaaaaaa60', '60bbbbbbbbbbbbbbbbb60'), (60, '60boobook60', '60aaaaaaaaaaaa60', '60bbbbbbbbbbbbbbbbb60')]
 
<_MainThread(MainThread, started 18744)>
 
<_MainThread(MainThread, started 18744)>
 
[(80, '80boobook80', '80aaaaaaaaaaaa80', '80bbbbbbbbbbbbbbbbb80'), (80, '80boobook80', '80aaaaaaaaaaaa80', '80bbbbbbbbbbbbbbbbb80'), (80, '80boobook80', '80aaaaaaaaaaaa80', '80bbbbbbbbbbbbbbbbb80')]
 
<_MainThread(MainThread, started 18744)>
 
[(90, '90boobook90', '90aaaaaaaaaaaa90', '90bbbbbbbbbbbbbbbbb90'), (90, '90boobook90', '90aaaaaaaaaaaa90', '90bbbbbbbbbbbbbbbbb90'), (90, '90boobook90', '90aaaaaaaaaaaa90', '90bbbbbbbbbbbbbbbbb90')]
 
<_MainThread(MainThread, started 18744)>
 
[(100, '100boobook100', '100aaaaaaaaaaaa100', '100bbbbbbbbbbbbbbbbb100'), (100, '100boobook100', '100aaaaaaaaaaaa100', '100bbbbbbbbbbbbbbbbb100'), (100, '100boobook100', '100aaaaaaaaaaaa100', '100bbbbbbbbbbbbbbbbb100')]
 
all over 
 
Wed Mar 28 11:08:34 2018 
 
Process finished with exit code 0 
 
 
这个时候是串行 花费了15秒 
 
 
多线程跑:
 
#coding=utf-8
 
import threading
 
import time
 
import cx_Oracle
 
from pprint import pprint
 
import csv
 
print time.asctime()
 
table_name = "dbtest.csv"
 
f = open(table_name + ".csv", "w") 
 
 
def query01():
 
  tname = threading.current_thread()
 
  aa=10
 
  conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
  cursor = conn.cursor()
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
  conn.close()
 
def query02():
 
  tname = threading.current_thread()
 
  aa=20
 
  conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
  cursor = conn.cursor()
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
  conn.close()
 
def query03():
 
  tname = threading.current_thread()
 
  aa=30
 
  conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
  cursor = conn.cursor()
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
  conn.close()
 
def query04():
 
  conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
  cursor = conn.cursor()
 
  tname = threading.current_thread()
 
  aa=40
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
  conn.close()
 
def query05():
 
  conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
  cursor = conn.cursor()
 
  tname = threading.current_thread()
 
  aa=50
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
  conn.close()
 
def query06():
 
  conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
  cursor = conn.cursor()
 
  tname = threading.current_thread()
 
  aa=60
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
  conn.close()
 
def query07():
 
  conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
  cursor = conn.cursor()
 
  tname = threading.current_thread()
 
  aa=70
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  return row
 
  cursor.close()
 
  conn.close()
 
def query08():
 
  conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
  cursor = conn.cursor()
 
  tname = threading.current_thread()
 
  aa=80
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
  conn.close()
 
def query09():
 
  conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
  cursor = conn.cursor()
 
  tname = threading.current_thread()
 
  aa=90
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
  conn.close()
 
def query10():
 
  conn = cx_Oracle.connect('tlcbuser/tlcbuser@20.5.101.31/tlyy')
 
  cursor = conn.cursor()
 
  tname = threading.current_thread()
 
  aa=100
 
  # cursor.execute("SELECT * FROM TEST100 WHERE ID = %s", [aa])
 
  cursor.prepare("""SELECT * FROM TEST100 WHERE ID=:id""")
 
  cursor.execute(None,{'id':aa})
 
  row = cursor.fetchall()
 
  print row
 
  return row
 
  cursor.close()
 
  conn.close() 
 
 
threads = []
 
t1 = threading.Thread(target=query01, name='query01')
 
threads.append(t1)
 
t2 = threading.Thread(target=query02, name='query02')
 
threads.append(t2)
 
t2 = threading.Thread(target=query03, name='query03')
 
threads.append(t2)
 
t2 = threading.Thread(target=query04, name='query04')
 
threads.append(t2)
 
t2 = threading.Thread(target=query05, name='query05')
 
threads.append(t2)
 
t2 = threading.Thread(target=query06, name='query06')
 
threads.append(t2)
 
t2 = threading.Thread(target=query07, name='query07')
 
threads.append(t2)
 
t2 = threading.Thread(target=query08, name='query08')
 
threads.append(t2)
 
t2 = threading.Thread(target=query09, name='query09')
 
threads.append(t2)
 
t2 = threading.Thread(target=query10, name='query10')
 
threads.append(t2) 
 
 
if __name__ == '__main__':
 
  for t in threads:
 
    # t.setDaemon(True)
 
    t.start()
 
    # t.run()
 
    # t.start()
 
    # print '3333333'
 
    print threading.current_thread()
 
    # print t.is_alive()
 
    # print '3333333'
 
  t.join()
 
  print "all over "
 
  print time.asctime() 
 
 
C:\Python27\python.exe C:/Users/tlcb/PycharmProjects/untitled/a2.py
 
Wed Mar 28 11:12:47 2018
 
<_MainThread(MainThread, started 22500)>
 
<_MainThread(MainThread, started 22500)>
 
<_MainThread(MainThread, started 22500)>
 
<_MainThread(MainThread, started 22500)>
 
<_MainThread(MainThread, started 22500)>
 
<_MainThread(MainThread, started 22500)>
 
<_MainThread(MainThread, started 22500)>
 
<_MainThread(MainThread, started 22500)>
 
<_MainThread(MainThread, started 22500)>
 
<_MainThread(MainThread, started 22500)>
 
[(40, '40boobook40', '40aaaaaaaaaaaa40', '40bbbbbbbbbbbbbbbbb40'), (40, '40boobook40', '40aaaaaaaaaaaa40', '40bbbbbbbbbbbbbbbbb40'), (40, '40boobook40', '40aaaaaaaaaaaa40', '40bbbbbbbbbbbbbbbbb40')]
 
[(60, '60boobook60', '60aaaaaaaaaaaa60', '60bbbbbbbbbbbbbbbbb60'), (60, '60boobook60', '60aaaaaaaaaaaa60', '60bbbbbbbbbbbbbbbbb60'), (60, '60boobook60', '60aaaaaaaaaaaa60', '60bbbbbbbbbbbbbbbbb60')]
 
[(80, '80boobook80', '80aaaaaaaaaaaa80', '80bbbbbbbbbbbbbbbbb80'), (80, '80boobook80', '80aaaaaaaaaaaa80', '80bbbbbbbbbbbbbbbbb80'), (80, '80boobook80', '80aaaaaaaaaaaa80', '80bbbbbbbbbbbbbbbbb80')]
 
[(50, '50boobook50', '50aaaaaaaaaaaa50', '50bbbbbbbbbbbbbbbbb50'), (50, '50boobook50', '50aaaaaaaaaaaa50', '50bbbbbbbbbbbbbbbbb50'), (50, '50boobook50', '50aaaaaaaaaaaa50', '50bbbbbbbbbbbbbbbbb50')]
 
[(10, '10boobook10', '10aaaaaaaaaaaa10', '10bbbbbbbbbbbbbbbbb10'), (10, '10sssssssss10', 'tlcb', 'tlcb'), (10, '10boobook10', '10aaaaaaaaaaaa10', '10bbbbbbbbbbbbbbbbb10')]
 
[(20, '20boobook20', '20aaaaaaaaaaaa20', '20bbbbbbbbbbbbbbbbb20'), (20, '20boobook20', '20aaaaaaaaaaaa20', '20bbbbbbbbbbbbbbbbb20'), (20, '20boobook20', '20aaaaaaaaaaaa20', '20bbbbbbbbbbbbbbbbb20')]
 
[(30, '30boobook30', '30aaaaaaaaaaaa30', '30bbbbbbbbbbbbbbbbb30'), (30, '30boobook30', '30aaaaaaaaaaaa30', '30bbbbbbbbbbbbbbbbb30'), (30, '30boobook30', '30aaaaaaaaaaaa30', '30bbbbbbbbbbbbbbbbb30')]
 
[(100, '100boobook100', '100aaaaaaaaaaaa100', '100bbbbbbbbbbbbbbbbb100'), (100, '100boobook100', '100aaaaaaaaaaaa100', '100bbbbbbbbbbbbbbbbb100'), (100, '100boobook100', '100aaaaaaaaaaaa100', '100bbbbbbbbbbbbbbbbb100')]
 
[(90, '90boobook90', '90aaaaaaaaaaaa90', '90bbbbbbbbbbbbbbbbb90'), (90, '90boobook90', '90aaaaaaaaaaaa90', '90bbbbbbbbbbbbbbbbb90'), (90, '90boobook90', '90aaaaaaaaaaaa90', '90bbbbbbbbbbbbbbbbb90')]
 
all over 
 
Wed Mar 28 11:12:55 2018 
 
Process finished with exit code 0 
 
此时花了 8秒

补充知识：Python日志定制Logstash JSON格式

近期完成了日志收集系统的调试工作，本文旨在分享一种优化Logstash处理Python日志的方案。

众所周知，Logstash使用Grok插件解析日志时会消耗大量CPU资源，因为正则匹配过程较为耗时。

基于Logstash的性能调优建议，我们可以预先将日志生成为JSON格式。针对Python程序，具体实现方式如下：

主要有两种实现路径：第一种是在生成JSON后，直接发送至Logstash监听端口；第二种是将JSON写入文件，利用Logstash的Tail插件读取文件，直接解析单行JSON日志数据。

Python内置的logging模块可以方便地调整为输出JSON格式。此外，已有社区开发的Python Logstash模块满足了这一需求。

Logstash JSON格式日志配置


import logging
import logstash
import sys

host = 'localhost'

test_logger = logging.getLogger('python-logstash-logger')
test_logger.setLevel(logging.INFO)
test_logger.addHandler(logstash.LogstashHandler(host, 5959, version=1))
# test_logger.addHandler(logstash.TCPLogstashHandler(host, 5959, version=1))

test_logger.error('python-logstash: test logstash error message.')
test_logger.info('python-logstash: test logstash info message.')
test_logger.warning('python-logstash: test logstash warning message.')

# add extra field to logstash message
extra = {
  'test_string': 'python version: ' + repr(sys.version_info),
  'test_boolean': True,
  'test_dict': {'a': 1, 'b': 'c'},
  'test_float': 1.23,
  'test_integer': 123,
  'test_list': [1, 2, '3'],
}
test_logger.info('python-logstash: test extra fields', extra=extra)

python-logstash库支持AMQP协议方案。

AMQP协议支持方案


import logging
import logstash

# AMQP parameters
host = 'localhost'
username = 'guest'
password= 'guest'
exchange = 'logstash.py'

# get a logger and set logging level
test_logger = logging.getLogger('python-logstash-logger')
test_logger.setLevel(logging.INFO)

# add the handler
test_logger.addHandler(logstash.AMQPLogstashHandler(version=1,
                          host=host,
                          durable=True,
                          username=username,
                          password=password,
                          exchange=exchange))

# log
test_logger.error('python-logstash: test logstash error message.')
test_logger.info('python-logstash: test logstash info message.')
test_logger.warning('python-logstash: test logstash warning message.')

try:
  1/0
except:
  test_logger.exception('python-logstash: test logstash exception with stack trace')

无论采用何种方式，最终生成的日志格式符合规范即可。

JSON日志输出格式示例


{
  "@source"=>"unknown",
  "@type"=>"nginx",
  "@tags"=>[],
  "@fields"=>{
    "remote_addr"=>"192.168.0.1",
    "remote_user"=>"-",
    "body_bytes_sent"=>"13988",
    "request_time"=>"0.122",
    "status"=>"200",
    "request"=>"GET /some/url HTTP/1.1",
    "request_method"=>"GET",
    "http_referrer"=>"http://www.example.org/some/url",
    "http_user_agent"=>"Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.1 (KHTML, like Gecko) Chrome/21.0.1180.79 Safari/537.1"
  },
  "@timestamp"=>"2012-08-23T10:49:14+02:00"
}

此处简要说明，该模块的实际使用体验并不理想。我在Python中将日志格式化为JSON字符串，原本期望在Elasticsearch中像Grok解析那样将日志拆解为独立字段，但实际情况是整条日志仍被包裹在message字段中。相信读者已理解我的意思，这导致在Kibana中进行搜索变得十分困难。

在Kibana搜索时，我常使用类似 source:xxx AND level:INFO 的查询语句。结果正如前述，整条日志内容都堆积在@message字段内，无法直接利用结构化字段进行高效检索。

以上关于Python串行与并行执行实例及日志优化的内容分享完毕，希望能为读者提供参考，欢迎继续支持本站。