wxsheng 发表于 2018-10-5 08:28:18

Python实现MySQL状态监控

  #! /usr/bin/python
  # -*- encoding: utf8 -*-
  #from __future__ import division
  import sys
  import MySQLdb
  # 数据库配置参数
  host = '192.168.20.235'
  user = 'root'
  password = 'root'
  db = 'orca_cmdb'
  #----------------------------------------------------------------------
  def getConn(host, user, passwd, db):
  try:
  conn = MySQLdb.connect(host=host, user=user, passwd=passwd, db=db)
  return conn
  except Exception, e:
  print "Error %d: %s" % (e.args, e.args)
  sys.exit(1)
  #----------------------------------------------------------------------
  def closeConn(conn):
  """关闭 mysql connection"""
  conn.close()
  #----------------------------------------------------------------------
  def getValue(conn, query):
  """ 查询相关参数 返回一个值 """
  cursor = conn.cursor()
  getNum=cursor.execute(query)
  if getNum>0:
  result = cursor.fetchone()
  else:
  result=['0']
  return int(result)
  def getQuery(conn, query):
  """ 查询相关参数 返回多个值 """
  cursor = conn.cursor()
  cursor.execute(query)
  result = cursor.fetchall()
  return result
  #执行查询的总次数
  Questions = "show global status like 'Questions'"
  #服务器已经运行的时间(以秒为单位)
  Uptime = "show global status like 'Uptime'"
  Com_commit = "show global status like 'Com_commit'"
  Com_rollback = "show global status like 'Com_rollback'"
  #从硬盘读取键的数据块的次数。如果Key_reads较大,则Key_buffer_size值可能太小。
  #可以用Key_reads/Key_read_requests计算缓存损失率
  Key_reads = "show global status like 'Key_reads'"
  #从缓存读键的数据块的请求数
  Key_read_requests = "show global status like 'Key_read_requests'"
  #向硬盘写入将键的数据块的物理写操作的次数
  Key_writes = "show global status like 'Key_writes'"
  #将键的数据块写入缓存的请求数
  Key_write_requests = "show global status like 'Key_write_requests'"
  #是否有innodb引擎,5.5版本后没有了该参数。所以有特殊处理
  Have_innodb = "show global variables like 'have_innodb'"
  #不能满足InnoDB必须单页读取的缓冲池中的逻辑读数量。
  Innodb_buffer_pool_reads = "show global status like 'Innodb_buffer_pool_reads'"
  #InnoDB已经完成的逻辑读请求数
  Innodb_buffer_pool_read_requests = "show global status like 'Innodb_buffer_pool_read_requests'"
  #查询缓存被访问的次数
  Qcache_hits = "show global status like 'Qcache_hits'"
  #加入到缓存的查询数量,缓存没有用到
  Qcache_inserts = "show global status like 'Qcache_inserts'"
  #当前打开的表的数量
  Open_tables = "show global status like 'Open_tables'"
  #已经打开的表的数量。如果Opened_tables较大,table_cache 值可能太小
  Opened_tables = "show global status like 'Opened_tables'"
  #创建用来处理连接的线程数。如果Threads_created较大,你可能要
  #增加thread_cache_size值。缓存访问率的计算方法Threads_created/Connections
  Threads_created = "show global status like 'Threads_created'"
  #试图连接到(不管是否成功)MySQL服务器的连接数。缓存访问率的计算方法Threads_created/Connections
  Connections = "show global status like 'Connections'"
  #Com_select/s:平均每秒select语句执行次数
  #Com_insert/s:平均每秒insert语句执行次数
  #Com_update/s:平均每秒update语句执行次数
  #Com_delete/s:平均每秒delete语句执行次数
  Com_select = "show global status like 'Com_select'"
  Com_insert = "show global status like 'Com_insert'"
  Com_update = "show global status like 'Com_update'"
  Com_delete = "show global status like 'Com_delete'"
  Com_replace = "show global status like 'Com_replace'"
  #不能立即获得的表的锁的次数。如果该值较高,并且有性能问题,你应首先优化查询,然后拆分表或使用复制。
  Table_locks_waited = "show global status like 'Table_locks_waited'"
  #立即获得的表的锁的次数
  Table_locks_immediate = "show global status like 'Table_locks_immediate'"
  #服务器执行语句时自动创建的内存中的临时表的数量。如果Created_tmp_disk_tables较大,
  #你可能要增加tmp_table_size值使临时 表基于内存而不基于硬盘
  Created_tmp_tables = "show global status like 'Created_tmp_tables'"
  #服务器执行语句时在硬盘上自动创建的临时表的数量
  Created_tmp_disk_tables = "show global status like 'Created_tmp_disk_tables'"
  #查询时间超过long_query_time秒的查询的个数 缓慢查询个数
  Slow_queries = "show global status like 'Slow_queries'"
  #没有主键(key)联合(Join)的执行。该值可能是零。这是捕获开发错误的好方法,因为一些这样的查询可能降低系统的性能。
  Select_full_join = "show global status like 'Select_full_join'"
  if __name__ == "__main__":
  conn = getConn(host, user, password, db)
  Questions = getValue(conn, Questions)
  Uptime = getValue(conn, Uptime)
  Com_commit = getValue(conn, Com_commit)
  Com_rollback = getValue(conn, Com_rollback)
  Key_reads = getValue(conn, Key_reads)
  Key_read_requests = getValue(conn, Key_read_requests)
  Key_writes = getValue(conn, Key_writes)
  Key_write_requests = getValue(conn, Key_write_requests)
  Qcache_hits = getValue(conn, Qcache_hits)
  Qcache_inserts = getValue(conn, Qcache_inserts)
  Open_tables = getValue(conn, Open_tables)
  Opened_tables = getValue(conn, Opened_tables)
  Threads_created = getValue(conn, Threads_created)
  Connections = getValue(conn, Connections)
  Com_select = getValue(conn, Com_select)
  Com_insert = getValue(conn, Com_insert)
  Com_update = getValue(conn, Com_update)
  Com_delete = getValue(conn, Com_delete)
  Com_replace = getValue(conn, Com_replace)
  Table_locks_immediate = getValue(conn, Table_locks_immediate)
  Table_locks_waited = getValue(conn, Table_locks_waited)
  Created_tmp_tables = getValue(conn, Created_tmp_tables)
  Created_tmp_disk_tables = getValue(conn, Created_tmp_disk_tables)
  Slow_queries = getValue(conn, Slow_queries)
  Select_full_join = getValue(conn, Select_full_join)
  print u"_____一般信息统计___________________"
  # QPS = Questions / Seconds
  QPS = str(round(Questions / Uptime, 5))
  print u"QPS (每秒查询次数): " + QPS
  TPS = str(round((Com_commit + Com_rollback)/Uptime, 5))
  print u"TPS(每秒执行的事务数量): " + TPS
  # Read/Writes Ratio
  rwr = str(round((Com_select + Qcache_hits)*1.0 / (Com_insert + Com_update + Com_delete + Com_replace) * 100, 5)) + "%"
  print u"Read/Writes占比: " + rwr + "\n"
  print u"_____缓存使用情况统计___________________"
  Key_buffer_read_hits = str(round((1 - Key_reads/Key_read_requests) * 100, 5)) + "%"
  Key_buffer_write_hits = str(round((1 - Key_writes/Key_write_requests) * 100, 5)) + "%"
  print u"MyISAM缓存读命中率(99.3% - 99.9%较好): " + str(Key_buffer_read_hits)
  print u"MyISAM缓存写命中率(99.3% - 99.9%较好): " + str(Key_buffer_write_hits)
  if Qcache_hits>0:
  Query_cache_hits = str(round(((Qcache_hits/(Qcache_hits + Qcache_inserts)) * 100), 5)) + "%"
  else:
  Query_cache_hits='0.0%'
  print u"Query Cache 命中率: " + Query_cache_hits
  cursor = conn.cursor()
  getFlag=cursor.execute(Have_innodb)
  if getFlag>0:
  result = cursor.fetchone()
  Have_innodb = result
  else:
  Have_innodb ="YES"
  if (Have_innodb == "YES"):
  Innodb_buffer_pool_reads = getValue(conn, Innodb_buffer_pool_reads)
  Innodb_buffer_pool_read_requests = getValue(conn, Innodb_buffer_pool_read_requests)
  # Innodb_buffer_read_hits = (1 - Innodb_buffer_pool_reads / Innodb_buffer_pool_read_requests) * 100%
  Innodb_buffer_read_hits = str(round((1 - Innodb_buffer_pool_reads/Innodb_buffer_pool_read_requests) * 100, 5)) + "%"
  print u"Innodb缓存命中率(建议96% - 99%): " + Innodb_buffer_read_hits
  Thread_cache_hits = str(round(((1 - Threads_created / Connections)) * 100, 5)) + "%"
  print u"线程缓存命中率=(应该90%以上): " + Thread_cache_hits + "\n"
  print u"_____慢查询sql监控________________"
  Slow_queries_per_second = str(round(Slow_queries*1.0 / (Uptime/60), 5))
  print "每分钟慢查询次数: " + Slow_queries_per_second
  Select_full_join_per_second = str(round(Select_full_join*1.0 / (Uptime/60), 5))
  print "每分钟无索引join操作: " + Select_full_join_per_second
  full_select_in_all_select = str(round((Select_full_join*1.0 / Com_select) * 100, 5)) + "%"
  print "无索引join操作占比: " + full_select_in_all_select
  lock_contention = str(round((Table_locks_waited*1.00 / Table_locks_immediate) * 100, 5)) + "%"
  print "MyISAM加锁等待比率(3% bad): " + lock_contention
  print "当期已打开表数: " + str(Open_tables)
  print "累计打开表数: " + str(Opened_tables)
  Temp_tables_to_disk = str(round((Created_tmp_disk_tables*1.0 / Created_tmp_tables) * 100, 5)) + "%"
  print u"临时表转化磁盘的比率:(最好不要超过10%) " + Temp_tables_to_disk
  closeConn(conn)

页: [1]
查看完整版本: Python实现MySQL状态监控