def run(self):
"""threadproc"""
self.hamon = Hamon()
self.grid = NZXTGrid()
print("Controller has started")
TIME_SLICE = 1000 # sampling period, msec
MAX_SLEEP = 250 # max sleep interval (sleep in smaller intervals, makes UI more responsive)
counter = 0
td = timediff()
while not self.shutdown:
td.reset()
self.dowork() # do all controller stuff, once per time slice
td.now()
exec_time = td.ms
remaining = TIME_SLICE
sleep_count = 0
while remaining > 0:
if self.shutdown or self.settingsTS < self.appsettings.timestamp:
break # early exit if needed
td.now()
remaining = TIME_SLICE - td.ms
if (remaining > MAX_SLEEP): remaining = MAX_SLEEP
if (remaining > 0):
time.sleep(remaining / 1000.0)
sleep_count += 1
td.now()
slice_time = td.ms
#print ("loop={0}, exec={1}, slice={2}, sleep count={3}".format(counter, exec_time, slice_time, sleep_count))
counter += 1
self.grid.close()
self.hamon.close()
print("Controller has stopped")
def tmp():
av = []
final_start = datetime.datetime(2012, 5, 1, 0, 0, 0)
start = final_start
end = datetime.datetime.now()
one_day = datetime.timedelta(days=1)
while start < end:
# 本月(start)的起止时间点
start_time, end_time = get_start_end_for_month(start)
alarm = SysAlarm.objects.filter(start_time__gte=start_time, start_time__lte=end_time)
# 本月总时间
all_min = timediff(start_time, end_time, 'minute')
# 初始: 本月可用时间=本月总时间
av_time = all_min
# 初始: 本月可用时间(因自身原因导致故障)=本月总时间
av_self_time = all_min
for a in alarm:
if a.type in c.self_alarm_type:
av_self_time = av_time - timediff(a.start_time, a.end_time, 'minute')
av_time = av_time - timediff(a.start_time, a.end_time, 'minute')
# 全部原因导致的系统可用率
av_percent = round(((av_time + 0.0000000000001) / all_min), 6)
# 自身原因导致的系统可用率
av_self_percent = round(((av_self_time + 0.0000000000001) / all_min), 6)
av_color = c.green if av_percent > 0.9999 else c.red
av_self_color = c.green if av_self_percent > 0.9999 else c.red
av_percent = to_percent(av_percent)
av_self_percent = to_percent(av_self_percent)
av.append({'month':start.strftime("%Y-%m-%d"), 'time':av_time, 'self_time':av_self_time, 'percent':av_percent, 'self_percent':av_self_percent,
'color':av_color, 'self_color':av_self_color})
start = end_time + one_day
return av
def sys_alarm(request, month='2012-05-01'):
t = loader.get_template('monitor/sys_alarm_table.html')
d = datetime.datetime.strptime(month, "%Y-%m-%d")
start_time, end_time = get_start_end_for_month(d)
a = SysAlarm.objects.filter(start_time__gte=start_time, start_time__lte=end_time)
alarm = []
if a:
for i in a:
alarm.append({'type':i.type,
'start_time':i.start_time.strftime('%Y-%m-%d %H:%M:%S'),
'end_time':i.end_time.strftime('%Y-%m-%d %H:%M:%S'),
'duration': timediff(i.start_time, i.end_time, 'minute'),
'reason':i.reason if i.reason else '',
'wiki_url':i.wiki_url if i.wiki_url else '',
'comments':i.comments if i.comments else ''})
response = HttpResponse(t.render(Context({'alarm':alarm})))
# 业务监控的系统可用率
now = datetime.datetime.now()
expire = (now + datetime.timedelta(hours=8))
response['Expires'] = expire.strftime('%a, %d %b %Y %H:%M:%S %Z')
return response
def read_alarm_job():
start_time = datetime.datetime.now() - datetime.timedelta(minutes=36)
read_failure_data = AppAvailableData.objects.filter(name='read', result=False, time__gte=start_time).values('time')
failure_count = len(read_failure_data)
if failure_count >= c.read_alarm_time:
try:
start_time = read_failure_data[0]['time']
end_time = read_failure_data[failure_count - 1]['time']
type = 'read_bookmark'
time = get_date_and_time()[1]
msg = 'read failure count:%s,%s' % (str(failure_count), time)
sms(mobile_list=c.mobile_list, message_post=msg)
# 获取上一次的报警信息
latest = SysAlarm.objects.filter(type=type).order_by('-gmt_create')
if latest:
latest = latest[0]
# start_time是本次报警的开始时间
# latest.end_time是上次报警的结束时间
# 所有在算时间差的时候, start_time应该是时间差的end, latest.end_time应该是时间差的start
tdiff = timediff(latest.end_time, start_time)
if (not latest) or start_time > latest.end_time and tdiff > 600:
# 一次新的故障
alarm = SysAlarm(type=type, start_time=start_time, end_time=end_time)
alarm.save()
else:
# 添加-报警的定时任务是每10min执行一次, 如果本次报警的时间和上次开始的时间<10min, 则认为是同一次故障
# 同一次故障的持续报警, 只需修改上次的结束时间即可
latest.end_time = end_time
latest.save()
else:
# 以前没有过这样类型的故障
alarm = SysAlarm(type=type, start_time=start_time, end_time=end_time)
alarm.save()
except Exception, e:
c.logger.error(e)
