the problem with balancing in swarm mode

Hi, my docker swarm consists of 2 nodes: 1 manager and 1 worker. But the services installed on the worker node do not work. When I go to traffic.**.site, the site works every other time. (it opens 1 time, does not open 2 times, opens 3 times, and so on). Tell me what my mistake is.

services:
  reverse-proxy:
    image: traefik:v3.0
    ports:
      - "80:80"
    deploy:
      mode: "global"
      placement:
        constraints:
          - "node.role == manager"
      labels:
        - "traefik.http.routers.mydashboard.rule=Host(`traefik.**.site`)"
        - "traefik.http.routers.mydashboard.service=api@internal"
        - "traefik.http.services.mydashboard.loadbalancer.server.port=80"
    volumes:
      - "/var/run/docker.sock:/var/run/docker.sock"
    command:
      - "--api.insecure=true"
      - "--log.level=DEBUG"
      - "--providers.swarm.endpoint=unix:///var/run/docker.sock"
    networks:
      - proxy
    
  whoami:
    image: "traefik/whoami"
    networks:
      - "proxy"
    deploy:
      mode: "global"
      labels:
        - "traefik.http.routers.whoami.rule=Host(`whoami.**.site`)"
        - "traefik.http.services.whoami-service.loadbalancer.server.port=80"

  web:
    image: nginx
    networks:
      - "proxy"
    deploy:
      mode: "global"
      labels:
        - "traefik.http.routers.webs.rule=Host(`web.**.site`)"
        - "traefik.http.services.webs-service.loadbalancer.server.port=80"

networks:
  proxy:
    name: proxy
    driver: overlay
    attachable: true

Is your overlay network working? Do you use a VLAN/vSwitch? Try a ping with payload > 1500 bytes between containers on different nodes.

Thanks for the advice, I'll try. I don't use VLAN/vSwitch.