moby--moby/docs/swarm/swarm-tutorial/rolling-update.md

Apply rolling updates to a service

In a previous step of the tutorial, you scaled the number of instances of a service. In this part of the tutorial, you deploy a service based on the Redis 3.0.6 container image. Then you upgrade the service to use the Redis 3.0.7 container image using rolling updates.

1. If you haven't already, open a terminal and ssh into the machine where you run your manager node. For example, the tutorial uses a machine named manager1.

2. Deploy Redis 3.0.6 to the swarm and configure the swarm with a 10 second update delay:

   $ docker service create \
     --replicas 3 \
     --name redis \
     --update-delay 10s \
     --update-parallelism 1 \
     redis:3.0.6
   
   0u6a4s31ybk7yw2wyvtikmu50
   

   You configure the rolling update policy at service deployment time.

   The --update-parallelism flag configures the number of service tasks that the scheduler can update simultaneously. When updates to individual tasks return a state of RUNNING or FAILED, the scheduler schedules another task to update until all tasks are updated.

   The --update-delay flag configures the time delay between updates to a service task or sets of tasks.

   You can describe the time T as a combination of the number of seconds Ts, minutes Tm, or hours Th. So 10m30s indicates a 10 minute 30 second delay.

3. Inspect the redis service:

   $ docker service inspect --pretty redis
   
   ID:             0u6a4s31ybk7yw2wyvtikmu50
   Name:           redis
   Mode:           Replicated
    Replicas:      3
   Placement:
    Strategy:	    Spread
   UpdateConfig:
    Parallelism:   1
    Delay:         10s
   ContainerSpec:
    Image:         redis:3.0.6
   Resources:
   

4. Now you can update the container image for redis. The swarm manager applies the update to nodes according to the UpdateConfig policy:

   $ docker service update --image redis:3.0.7 redis
   redis
   

   The scheduler applies rolling updates as follows:

   • Stop the initial number of tasks according to --update-parallelism.
   • Schedule updates for the stopped tasks.
   • Start the containers for the updated tasks.
   • After an update to a task completes, wait for the specified delay period before stopping the next task.

5. Run docker service inspect --pretty redis to see the new image in the desired state:

   $ docker service inspect --pretty redis
   
   ID:             0u6a4s31ybk7yw2wyvtikmu50
   Name:           redis
   Mode:           Replicated
    Replicas:      3
   Placement:
    Strategy:	    Spread
   UpdateConfig:
    Parallelism:   1
    Delay:         10s
   ContainerSpec:
    Image:         redis:3.0.7
   Resources:
   

6. Run docker service tasks <TASK-ID> to watch the rolling update:

   $ docker service tasks redis
   
   ID                         NAME     SERVICE  IMAGE        LAST STATE              DESIRED STATE  NODE
   dos1zffgeofhagnve8w864fco  redis.1  redis    redis:3.0.7  Running 37 seconds      Running        worker1
   9l3i4j85517skba5o7tn5m8g0  redis.2  redis    redis:3.0.7  Running About a minute  Running        worker2
   egiuiqpzrdbxks3wxgn8qib1g  redis.3  redis    redis:3.0.7  Running 48 seconds      Running        worker1
   

   Before Swarm updates all of the tasks, you can see that some are running redis:3.0.6 while others are running redis:3.0.7. The output above shows the state once the rolling updates are done.

Next, learn about how to drain a node in the Swarm.