value processing in each process in Python Pool()

Question

I want to get the difference between 2 values of many data sources in real-time, so I am using Pool() to parallelly run functions infinitely. But the problem is, that the "next" process can't get the previous data to calculate stuff because I declare global variables in the function.

def initializer():
    global height
    global block_time_2
    height = 0
    block_time_2 = 0

def query(name, rpc):
    block_time_1 = 0 #the first value
    global height #get the data from RPC. This is the cause of the problem, always get 0
    try:
        request = Request("{}/block".format(rpc), None, header)
        data = json.loads(urlopen(request).read())
        if height != int(data['result']['block']['header']['height']):
            height = int(data['result']['block']['header']['height'])
            block_time_1 = parser.parse(data['result']['block']['header']['time']) # first variables is updated
            blockTime = getBlockTime(name, block_time_1, height, rpc) # get the second value and calculate the difference
            return blockTime
        else:
            return 
    except:
        traceback.print_exc()

def getBlockTime(name, block_time_1, height, rpc):
# second value. Pool doesn't share the memory among processes so doesn't need to
# declare an array for different data source. After calculation, this will be replace
# with the first value, so the fetch function should be called fewer time.
    global block_time_2
    print("{} time block 2: {}".format(name, block_time_2)) # always get 0, but sometimes randomly get the data
    if block_time_2 != 0:
        blockTime = (block_time_1 - block_time_2).total_seconds()
        block_time_2 = block_time_1
        return blockTime
    else:
        print("{} block time 2  recalculated".format(name))
        request = Request("{}/block?height={}".format(rpc, height - 1), None, header)
        data = json.loads(urlopen(request).read())
        block_time_2 = parser.parse(data['result']['block']['header']['time'])
        blockTime = (block_time_1 - block_time_2).total_seconds()
        block_time_2 = block_time_1
        return blockTime

def multi_query(args):
    return query(*args)

if __name__ == '__main__':
# try many ways but not help
#    manager = Manager()
#    chainsData = manager.dict()
#    q = Queue()
    p = Pool(25, initializer, ())
    data = json.load(
        open('/home/source-data.json'))
    while True:
        chainsData = dict(p.map(multi_query, [
            (data["source-1"]["name"],   data["source-1"]["rpc"]),
            (data["source-2"]["name"],   data["source-2"]["rpc"]),
            (data["source-3"]["name"],   data["source-3"]["rpc"]),
        ]))
        export(chainsData) # the function just works, the only problem is it fetched from RPC so many times
        print("done\n---------")
        time.sleep(1)

Manager() and other functions as far as I learned are about sharing value among processes, but I can't find any ways to get the process using data of the previous process (in infinite while True loop). I know that the global variables I declared above are the problem, but I can't find any solutions for it yet.