python 历史数据删除法则
#!/usr/bin/env python
#-*- coding:UTF-8 -*-
"""
@Item : database Backup
@Author : Villiam Sheng
@Date : 2012-08-13
@Funtion:
"""
import os,sys,time,re,socket,threading,json,base64,traceback,ConfigParser,fcntl,struct
from rsync_log import rsync_log
from rsync_post import rsync_post
from statvfs import F_BLOCKS,F_BAVAIL,F_BSIZE
pcg = 0
""" 生成斐波那契数列"""
lists = []
a,b = 0,1
while b <= 365:
a,b = b ,a+b
lists.append(b)
class rsync_thread(threading.Thread):
def __init__(self,path):
threading.Thread.__init__(self)
self.log = rsync_log()
self.path = path
""" 计算当前磁盘的使用百分比"""
def disk(self):
try:
vfs = os.statvfs(self.path)
disk_full = int(vfs[F_BLOCKS]*vfs[F_BSIZE]/1024/1024/1024)
disk_free = int(vfs[F_BAVAIL]*vfs[F_BSIZE]/1024/1024/1024)
return '%.1f'%(float(disk_full - disk_free) / disk_full * 100)
except:
self.log.log_info('rsync_info.err','dfile.disk',traceback.print_exc())
return traceback.print_exc()
def run(self):
global pcg
old_df = [] # 上一年的删除历史文件
new_df = [] # 今年的删除历史文件
sf = [] # 保留的历史文件
res = {} # 所有文件的天数及文件的路径
rs = 0 # 删除文件的总和
size = [] # 获取删除文件的大小
msize = [] # 今天备份所有文件的大小
tday_size = [] # 今天备份文件的大小
ms = 0 # 今天备份文件的总和
year = time.localtime().tm_year
""" 得到文件的天数，以文件名作为key，天数作为value """
for root,dirs,files in os.walk(self.path):
for fpath in files:
res[os.path.join(root,fpath)] = time.localtime(os.stat(os.path.join(root,fpath)).st_ctime).tm_yday
""" 判断文件的天数是否符合斐波那契数列，符合条件append到sf列表中,不符合append df列表中 """
for v,k in res.items():
if k in lists:
sf.append(k)
self.log.log_info('log_info.save','dfile.rsync_thread','%s:::%s'%(v,k))
elif k not in lists:
if year != time.localtime(os.stat(v).st_ctime).tm_year:
old_df.append({v:k})
else:
new_df.append({v:k})
try:
for s in range(len(new_df)):
for f,k in new_df[s].items():
tday_size.append(k)
if max({}.fromkeys(tday_size).keys()) == k:
msize.append(os.path.getsize(f))
except:
pass
c = []
pcg = float(self.disk())
""" 判断今天是否有新的文件备份，在删除的列表中删除最后一天的数据，但必须保证磁盘的使用百分比大于等于 %91 """
if time.localtime().tm_yday in res.values():
if len(old_df) != 0:
for s in range(len(old_df)):
for f,k in old_df[s].items():
c.append(k)
for s in range(len(old_df)):
for f,k in old_df[s].items():
if min({}.fromkeys(c).keys()) == k and pcg > 91:
size.append(os.path.getsize(f))
os.system('rm -frv %s' %f)
self.log.log_info('log_info.delete','remove cmd','rm -frv %s %s'%(f,k))
elif pcg <= 91:
break
pcg = float(self.disk())
elif len(new_df) != 0:
for s in range(len(new_df)):
for f,k in new_df[s].items():
c.append(k)
for s in range(len(new_df)):
for f,k in new_df[s].items():
if min({}.fromkeys(c).keys()) == k and pcg > 91:
size.append(os.path.getsize(f))
os.system('rm -frv %s' %f)
self.log.log_info('log_info.delete','remove cmd','rm -frv %s %s'%(f,k))
elif pcg <= 91:
break
pcg = float(self.disk())
for s in size:
rs += s
for m in msize:
ms += m
self.log.log_info('log_info.delete','Disk release %s %s MB'%(self.path,rs /1024/1024) ,'Disk append %s %s MB'%(self.path,ms /1024/1024))
else:
self.log.log_info('log_info.delete','Disk files ',' %s No update file' %self.path)
sys.exit()
class rsync_dfile(object):
def __init__(self):
def work(self):
fp = open('/proc/mounts','r')
m_info=fp.readlines()
fp.close()
data = {}
sections = []
for i in m_info:
if i.find('data=ordered') != -1 or i.find('mfs') != -1 or i.find('nfs') != -1:
if os.path.ismount(str(i.split()[1])):
if str(i.split()[1]) != '/':
if str(i.split()[1]) != '/root':
if str(i.split()[1]) != '/var':
if len(i.split()[1]) != 1:
if not i.find('sunrpc') != -1:
rs_thread = rsync_thread(i.split()[1])
rs_thread.start()
if __name__ == "__main__":
rs = rsync_dfile()
while True:
rs.work()
if pcg <= 91:
break
本文出自 “欢迎评论，欢迎点赞” 博客，谢绝转载！