pyads通知模块无异常抛出但程序停滞的问题求助

我开发了一款信号处理工具，正常运行状态符合预期，仅当出现可处理异常时程序才会终止。我已使用try语句捕获并记录所有可能出现的异常，但目前程序出现了无异常抛出却直接停滞的情况。我原本考虑实现一个超时定时器，当任务执行超时触发异常，但或许有更简便的方案？

经排查发现，当ADS服务器断开连接时，pyads的通知模块并未抛出异常。

import pyads
import ctypes
from time import sleep
import numpy as np
import pandas as pd
import logging
from os import getcwd

class VibrationData:
    
    def __init__(self):
        self.plc            = pyads.Connection('127.0.0.1.1.1', pyads.PORT_TC3PLC1, ip_address='127.0.1.1')
        self.plc.open()
        self.n_channels     = self.plc.read_by_name("GVL_Constants.nMaxChannels", plc_datatype=pyads.PLCTYPE_UDINT)
        self.signal_length  = self.plc.read_by_name("GVL_Constants.nSignalLength", plc_datatype=pyads.PLCTYPE_UDINT)
        self.sample_rate    = self.plc.read_by_name("GVL_Constants.fSampleRate", plc_datatype=pyads.PLCTYPE_LREAL)
        c_Measurement_Range_Acc = self.plc.read_by_name("GVL_Constants.fMeasurementRangeAcc", plc_datatype=pyads.PLCTYPE_LREAL)
        c_Resolution        = self.plc.read_by_name("GVL_Constants.fResolutionAcc", plc_datatype=pyads.PLCTYPE_LREAL)
        c_Gravitational_Acc = 9.81
        c_Sensitivity_Acc   = 0.1
        self.scale          = c_Measurement_Range_Acc*c_Gravitational_Acc/(c_Sensitivity_Acc*c_Resolution)
        self.buffer_factor  = self.plc.read_by_name("GVL_Constants.nBufferFactor", plc_datatype=pyads.PLCTYPE_UDINT)
        self.buffer_length  = int(self.signal_length/self.buffer_factor)
        self.sample_idx     = 0

    def get_structsize(self):
        raw_acc_tuple = list()
        for i in range(self.n_channels):
            raw_acc_tuple.append(tuple(('Iepe'+str(i),pyads.PLCTYPE_DINT,self.buffer_length)))
        self.structure = tuple(raw_acc_tuple)
        self.size_of_struct = int(pyads.size_of_structure(self.structure))
        self.attr = pyads.NotificationAttrib(self.size_of_struct)
        
    def scale_raw_data(self, raw_data):
        for i in range(self.n_channels):
            ch_name = 'Iepe'+str(i)
            self.raw_acc[i,int((self.sample_idx)*self.buffer_length):int((self.sample_idx+1)*self.buffer_length)] = raw_data[ch_name]
#initializing of the exception logger
logging.basicConfig(format='%(asctime)s %(levelname)s %(name)s %(message)s', filename='errorlog.txt', level=logging.DEBUG)
#initialize control variables for signal processing

init_state = True

while True:
    try:
        if init_state == True:
            init_state = False
            x = VibrationData()
            data_arrived = False
            acquisition_active = False
            sample_idx = 1
            
            @x.plc.notification(ctypes.c_ubyte * x.size_of_struct)
            def callback(handle, name, timestamp, value):
                global data_arrived
                values = pyads.dict_from_bytes(value, x.structure)
                x.scale_raw_data(values)
                x.sample_idx += 1
                if x.sample_idx == x.signal_length/x.buffer_length:
                    data_arrived = True
                    x.sample_idx = 1          
            #print("Init done!")
        if acquisition_active == False:  
            notification = x.plc.add_device_notification('MAIN_IEPE.anBufferPython',x.attr,callback)
            acquisition_active = True
        if data_arrived == True:
            print('data arrived!')
            acquisition_active = False
            data_arrived = False
            x.plc.del_device_notification(notification[0],notification[1])
    except:
        logging.exception('Exception occured.')
        init_state = True
        x.plc.del_device_notification(notification[0],notification[1])
        sleep(5)

内容的提问来源于stack exchange，提问作者Maik_FLS