Communication has a big latency

[astatio]

Waiting for a server's response seems to normally take around ~5000ms.

How to reproduce:

I have tested this with the following scenario:

Client side (Kotlin(JVM)):

import co.gongzh.procbridge.Client

val procClient = Client(
    "0.0.0.0",
    8000
)

    val start = System.currentTimeMillis()
    procClient.request("ping", null)
    println("Latency : ${System.currentTimeMillis() - start}ms")

Server side (Python):

import procbridge as pb


def handler(method, args):
    if method == "initialtest1":
        return "Initial test 1 triggered"
    if method == "initialtest2":
        return "Initial test 2 triggered"
    if method == "echo":
        return args
    if method == "ping":
        return "pong"


if __name__ == '__main__':
    port = 8000 
    s = pb.Server('0.0.0.0', port, handler)
    s.start(daemon=False)
    print("Procbridge server is on {}...".format(port))

Run it several time or put it in a loop. You should get a similar result:

Latency : 5020ms
Latency : 1ms
Latency : 5001ms
Latency : 5009ms
Latency : 5003ms
Latency : 5004ms
Latency : 5004ms
Latency : 5003ms
Latency : 5004ms
Latency : 3ms
Latency : 2ms
Latency : 5008ms
Latency : 1ms
Latency : 5ms
Latency : 5008ms
Latency : 1ms
Latency : 1ms
Latency : 2ms
Latency : 5010ms
Latency : 5004ms
Latency : 5003ms
Latency : 5004ms
Latency : 5006ms
Latency : 5002ms

[gongzhang]

Hi, @astatio, thanks for the detailed description.

I took a quick look and it appears that something blocked and timed out at the 5 second mark. But I don't have a development environment at hand right now, can you help to confirm whether the blocking occurs on the server side or the client side? For example, you can set a breakpoint at procbridge/__init__.py#L126 and procbridge/__init__.py#L135 in the Python server-side program. The former is the location where the client request is received, and the latter is the location where the response data is sent.