Discarded report from unrecognized worker

[michie1]

Prerequisites

• I am using the latest version of Locust
• I am reporting a bug, not asking a question

Description

This is a follow up of #3000. The UI now shows the correct number, but only when one worker node is used. When adding --processes 10 and specifying 10 users in the UI I still get a lower number in the UI compared with what the nodejs command is showing.

The logs is saying "locust.runners: ... reported that is has stopped" and later "Discarded report from unrecognized worker ...". This is logged multiple times for multiples runners.

const http = require('http');

let requestCount = 1;

const server = http.createServer((req, res) => {
  setTimeout(() => {
    console.log(requestCount);
    res.writeHead(200, { 'Content-Type': 'text/plain' });
    res.end(`Request count: ${requestCount}`);
    requestCount++; // Increment the request counter
  }, 1000);
});

const PORT = 3000;
server.listen(PORT, () => {
    console.log(`Server running at http://localhost:${PORT}`);
});

Command line

locust -s 10 --processes 10

Locustfile contents

from locust import HttpUser, task

i = 0

class HelloWorldUser(HttpUser):
    @task
    def hello_world(self):
        global i
        print("before", i)
        self.client.get("/")
        print("after", i)
        i += 1

### Python version

3.10

### Locust version

2.32.4

### Operating system

Ubuntu 22.04.5 WSL

[michie1]

[michie1]

Same holds when I remove the setTimeout (I thought maybe that would introduce count differences) and specify the Runtime to 10 seconds instead of clicking on the "Stop" button.

(--headless returns the correct count.)

[cyberw]

I can reproduce this as well. I wonder when it broke... Anyways, I'm on it...

[cyberw]

Actually, I'm only able to reproduce it with --stop-timeout. I still sometimes get the log message at shutdown without -s, but then no requests are actually missing (so probablly going to drop that log message level to debug). So still a bug, but much smaller scope...