GitHub - sudo-bmitch/docker-stack-wait: Wait for a docker stack deploy to complete

Waits for a docker stack deploy to complete.

Archive Notice

This project has been archived and is no longer being maintained. Docker has added the --detach CLI flag which allows similar functionality to be implemented with:

docker stack deploy --detach=false -c docker-compose.yaml $stack_name

CLI Usage

Example Usage:

docker-stack-wait.sh $stack_name

Help output:

$ ./docker-stack-wait.sh -h
docker-stack-wait.sh [opts] stack_name
  -f filter: only wait for services matching filter, may be passed multiple
             times, see docker stack services for the filter syntax
  -h:        this help message
  -l flags:  Print logs of relevant services at end.
             Flags are passed directly to the end of 'docker service logs'.
             Example usage: -l '--tail 20' or -l '--since 20m'
  -n name:   only wait for specific service names, overrides any filters,
             may be passed multiple times, do not include the stack name prefix
  -r:        treat a rollback as successful
  -s sec:    frequency to poll service state (default 5 sec)
  -t sec:    timeout to stop waiting (default 3600 sec)

Usage as container

An image is available at:

Docker Hub: sudobmitch/docker-stack-wait
GHCR: ghcr.io/sudo-bmitch/docker-stack-wait

To use this image, you will need to mount the docker socket:

$ docker run --rm -it \
  -v /var/run/docker.sock:/var/run/docker.sock \
  sudobmitch/docker-stack-wait $stack_name

or with an alias

$ alias docker-stack-wait='docker run --rm -it \
  -v /var/run/docker.sock:/var/run/docker.sock \
  sudobmitch/docker-stack-wait'

Development

To test changes to the script easily, you can use the example example-docker-compose.yml file with:

docker-compose -f dind-docker-compose.yml up
docker-compose -f dind-docker-compose.yml exec dind sh
docker node ls || docker swarm init
docker stack deploy --compose-file work/example-docker-compose.yml the_stack
./work/docker-stack-wait.sh the_stack

Filter Examples

The -n and -f options allow to select a subset of the services in a stack. With the following compose yml file:

version: '3.7'

services:
  normal:
    image: busybox
    command: /bin/sh -c ":>/healthy; tail -f /dev/null"
    deploy:
      labels:
        deploy.wait: "true"
        deploy.quick: "true"
    healthcheck:
      test: /bin/sh -c "[ -f /healthy ] && exit 0 || exit 1"
      interval: 15s
      start_period: 60s
      retries: 3

  slow:
    image: busybox
    command: /bin/sh -c "sleep 50; :>/healthy; tail -f /dev/null"
    deploy:
      labels:
        deploy.wait: "true"
    healthcheck:
      test: /bin/sh -c "[ -f /healthy ] && exit 0 || exit 1"
      interval: 15s
      start_period: 60s
      retries: 3

  tooslow:
    image: busybox
    command: /bin/sh -c "sleep 300; :>/healthy; tail -f /dev/null"
    deploy:
      labels:
        deploy.wait: "false"
    healthcheck:
      test: /bin/sh -c "[ -f /healthy ] && exit 0 || exit 1"
      interval: 15s
      start_period: 60s
      retries: 3

We can wait for only the first two services using labels:

docker-stack-wait.sh -f label=deploy.wait=true waittest

Or by waiting on individual service names:

docker-stack-wait.sh -n normal -n slow waittest

If you deploy a stack using multiple compose files, you can wait for the services in a single compose file using the following example that uses docker-compose to generate a list of services from one file:

wait_args=""
for arg in $(docker-compose -f docker-compose.yml config --services 2>/dev/null); do
  wait_args="${wait_args:+${wait_args} }-n $arg"
done
docker-stack-wait.sh $wait_args waittest