Rancher cli update service and skip update the load balancer

Rancher 1.x
1

I’m using the rancher cli to update the service to a new version.
While the upate process the rancher is creating a new load balancer and kill the old one. there is any way to avoid this and just update the service image and to use the exist load balancer without recreate a new one?

my docker compose file:

myapp:
  labels:
    io.rancher.container.pull_image: always
  image: image_path
myapp-lb:
  ports:
  - 80:7000
  labels:
    io.rancher.scheduler.global: 'true'
  tty: true
  image: rancher/load-balancer-service
  links:
  - myapp:myapp
  stdin_open: true

my rancher compose file:

myapp:
  scale: 2
  health_check:
    port: 7000
    interval: 2000
    initializing_timeout: 60000
    unhealthy_threshold: 3
    strategy: recreate
    upgrade_strategy:
     start_first: true
    request_line: GET "/healthcheck" "HTTP/1.0"
    healthy_threshold: 2
    response_timeout: 2000
myapp-lb:
  scale: 2
  load_balancer_config:
    haproxy_config: {}
  health_check:
    port: 42
    interval: 2000
    unhealthy_threshold: 3
    healthy_threshold: 2
    response_timeout: 2000

any idea?

2

@dsaydon have you tried removing the myapp-lb section when you upgrade your stack? Can you give that a go and see if that works?

3

it is working thanks!