检查resin的gc相关log的脚本(nagios插件版)-python

输入:log_path-日志位置;间隔监控时间(单位为分钟);full_gc_count-fullgc次数报警值;gc_interval-gc间隔时间报警值;gc_count-gc次数报警值

输出:当被检查log中有任一项报警内容大于报警值的时候，脚本退出状态为2，将所需报警内容打印在stdout中。

稍作修改也可以写在crontab里，作为定时任务执行。

脚本检测方法：

./check_gc.py -p/home/gc.log -f5 -g50 -i3 -t5

使用./
check_gc.py可以看到关于输入简要的帮助文档

#! /usr/local/bin/python3

import datetime
import re
import sys
import optparse

def gen_time(minutedelta):
format='%Y-%m-%dT%H:%M:%S'
return (datetime.datetime.today()-datetime.timedelta(minutes=minutedelta)).strftime(format)

def sub_times(time1,time2):
format='%Y-%m-%dT%H:%M:%S'
translate=datetime.datetime.strptime
return (translate(time1,format)-translate(time2,format)).total_seconds()

def change_to_data(time1):
format='%Y-%m-%dT%H:%M:%S'
return datetime.datetime.strptime(time1,format)

def check_log(log_path,examine_time):
report_dir={}
full_gc_re=re.compile('\[Full GC')
full_gc_count=0
gc_count=0
last_line_timestamp=0
min_gc_interval=999999
with open(log_path) as file:
for line in file:
line=line.strip()
time_line=line.split('.')[0]
#这里加入转换日期错误的捕获，保证之后所比较的都是期望的时间格式
try:
change_to_data(time_line)
except ValueError:
pass
else:
if time_line >=examine_time:
if last_line_timestamp != 0:
try:
gc_interval=sub_times(time_line,last_line_timestamp)
if gc_interval < min_gc_interval and gc_interval != 0:
min_gc_interval=gc_interval
except ValueError:
pass
else:
last_line_timestamp=time_line
else:
last_line_timestamp=time_line
gc_count+=1
if full_gc_re.search(line):
full_gc_count+=1
report_dir['full_gc']=full_gc_count
report_dir['min_gc_interval']=min_gc_interval
report_dir['gc_count']=gc_count
return report_dir

def check_result(report,full_gc,min_gc_interval,gc_count):
#这里只是简要的报警规则，报警规则可以更加的复杂
if full_gc <= report['full_gc']:
display(report)
sys.exit(2)
if min_gc_interval >= report['min_gc_interval']:
display(report)
sys.exit(2)
if gc_count <= report['gc_count']:
display(report)
sys.exit(2)

def display(display_data):
display_lines=''
for key in sorted(display_data):
display_lines+=(key+':'+str(display_data[key])+';')
print(display_lines)

def exec_check(begin_time,log_path,full_gc,min_gc_interval,gc_count):
time_begin=gen_time(begin_time)
result_dir=check_log(options.log_path,time_begin)
check_result(result_dir,full_gc,min_gc_interval,gc_count)

def generate_arguments():
#建议使用argparser，这里为了兼容旧版使用了optparse
parser=optparse.OptionParser()
parser.add_option('-p',action='store',type='string',dest='log_path',help='gc_log_path')
parser.add_option('-f',action='store',type='int',dest='full_gc',help='full_gc apper times')
parser.add_option('-g',action='store',type='int',dest='gc_count',help='gc apper times')
parser.add_option('-i',action='store',type='int',dest='min_gc_interval',help='the interval of gc')
parser.add_option('-t',action='store',type='int',dest='begin_time',help='check log from this value to now')
(options,others)=parser.parse_args()
return options

if __name__ == '__main__':
options=generate_arguments()
exec_check(options.begin_time,options.log_path,options.full_gc,options.min_gc_interval,options.gc_count)

附：gc日志格式：
2013-03-21T16:03:20.036+0800: 46930.500: [GC [PSYoungGen: 987599K->9744K(1013568K)] 1755375K->780204K(1800000K), 0.0511650 secs] [Times: user=0.12 sys=0.00, real=0.05 secs]
2013-03-21T16:03:39.397+0800: 46949.860: [GC [PSYoungGen: 987856K->11887K(1013120K)] 1758316K->784317K(1799552K), 0.0581740 secs] [Times: user=0.11 sys=0.01, real=0.06 secs]
2013-03-21T16:03:39.455+0800: 46949.919: [Full GC [PSYoungGen: 11887K->0K(1013120K)] [PSOldGen: 772429K->495896K(786432K)] 784317K->495896K(1799552K) [PSPermGen: 157441K->157441K(170560K)], 2.6535560 secs] [Times: user=2.66 sys=0.00, real=2.66 secs]
2013-03-21T16:04:28.456+0800: 46998.919: [GC [PSYoungGen: 978112K->34038K(998016K)] 1474008K->539300K(1784448K), 0.0918470 secs] [Times: user=0.24 sys=0.00, real=0.09 secs]
2013-03-21T16:04:49.390+0800: 47019.854: [GC [PSYoungGen: 998006K->11664K(1006272K)] 1503268K->541030K(1792704K), 0.0869980 secs] [Times: user=0.23 sys=0.00, real=0.08 secs]