将chrome实例与码头化的express应用程序连接时出错

我正在尝试在一个码头化的快递应用程序中通过一个chrome实例获得灯塔分数，但得到了：

flare_backend   | Error: connect ECONNREFUSED 127.0.0.1:9222
flare_backend   |     at __node_internal_captureLargerStackTrace (node:internal/errors:484:5)
flare_backend   |     at __node_internal_exceptionWithHostPort (node:internal/errors:662:12)
flare_backend   |     at TCPConnectWrap.afterConnect [as oncomplete] (node:net:1300:16) {
flare_backend   |   errno: -111,
flare_backend   |   code: 'ECONNREFUSED',
flare_backend   |   syscall: 'connect',
flare_backend   |   address: '127.0.0.1',
flare_backend   |   port: 9222
flare_backend   | }

我基本上有两种服务：

class TestService {
    async launchChrome () {
        return await launch({
            chromeFlags: ["--headless", "--no-sandbox", "--disable-gpu"],
            port: 9222
        });
    }

    async runLighthouse (url: string, chrome: LaunchedChrome, config: unknown) {
        return await lighthouse(url, { port: chrome.port }, config);
    }
}

async getTest () {
    const chrome = await TestService.launchChrome();

    console.log(chrome.port);

    const lighthouse = await TestService.runLighthouse(
        "https://www.google.com",
        chrome,
        { extends: "lighthouse:default" }
    );

    return lighthouse;
}

我的后端Dockerfile和docker compose文件如下所示：

ARG NODE_VERSION="18.12.1"
ARG ALPINE_VERSION="3.17"

FROM --platform=linux/amd64 node:${NODE_VERSION}-alpine${ALPINE_VERSION} AS base

WORKDIR /home/flare
RUN chown -R node:node /home/flare
RUN apk update && apk upgrade && \
    apk add --no-cache chromium && \
    rm -rf /var/cache/apk/*

ENV CHROME_BIN=/usr/bin/chromium-browser

COPY --chown=node:node package*.json ./
COPY --chown=node:node prisma ./prisma/

RUN npm install && npm cache clean --force
RUN npx prisma generate

COPY --chown=node:node . .

EXPOSE 8080

USER node

version: '3.9'
services:
    postgres:
        container_name: flare_postgres
        image: postgres:15.1-alpine
        restart: unless-stopped
        volumes:
            - ./backend/sql:/var/lib/postgresql/data
        env_file:
            - ./backend/.env
        ports:
            - 5432:5432
        networks:
            - flare_network
    redis:
        container_name: flare_redis
        image: redis:7.0.5-alpine
        restart: unless-stopped
        ports:
            - 6379:6379
        networks:
            - flare_network
    nginx:
        container_name: flare_nginx
        build: ./backend/nginx
        restart: unless-stopped
        ports:
            - 80:80
            - 443:443
        depends_on:
            - backend
            - frontend
        networks:
            - flare_network
    prisma:
        container_name: flare_prisma
        command: sh -c "prisma migrate deploy && prisma studio"
        build: ./backend/prisma
        restart: unless-stopped
        ports:
            - 5555:5555
        env_file:
            - ./backend/.env
        depends_on:
            - postgres
        networks:
            - flare_network
    backend:
        container_name: flare_backend
        command: sh -c "npm install && npm run dev"
        build: ./backend
        volumes:
            - ./backend:/home/flare
            - /home/flare/node_modules
        working_dir: /home/flare
        restart: unless-stopped
        depends_on:
            - postgres
            - redis
        ports:
            - 8080:8080
        networks:
            - flare_network
    frontend:
        container_name: flare_frontend
        command: sh -c "npm install && npm run dev"
        build: ./frontend
        volumes:
            - ./frontend:/home/flare
            - /home/flare/node_modules
        working_dir: /home/flare
        restart: unless-stopped
        depends_on:
            - backend
        ports:
            - 3000:3000
            - 24678:24678
        networks:
            - flare_network
networks:
    flare_network:
        driver: bridge

我试图更改一些配置等，但无法解决问题。当我尝试在本地运行此应用程序时，它运行时没有问题，但在码头化环境中没有问题。我怀疑这可能是网络问题，但也不应该是真正的问题，因为express应用程序和chrome实例位于同一个容器中。这里可能有什么问题？