Docker

The docker trigger lets you replace existing containers with their updated versions.

The trigger will:

• Clone the existing container specification
• Pull the new image
• Stop the existing container
• Remove the existing container
• Create the new container
• Start the new container (if the previous one was running)
• Remove the previous image (optionally)

Variables

Image Backup & Rollback

Every time a container is updated, Drydock automatically backs up the previous image metadata. Backups are retained per container up to the BACKUPCOUNT limit (default: 3), with older backups pruned automatically.

You can roll back to any retained backup via the API:

POST /api/backup/{id}/rollback

Optionally pass a specific backup ID in the request body:

{ "backupId": "uuid" }

If omitted, the most recent backup is used. Rollback pulls the backup image, stops the current container, recreates it with the backup image, and starts it.

Auto-Rollback

Drydock can automatically roll back a container update if the container becomes unhealthy. This requires the container to have a Docker HEALTHCHECK defined.

Configure auto-rollback using container labels:

After a successful update, if dd.rollback.auto is enabled, Drydock polls the container health status at the configured interval. If the container becomes unhealthy within the monitoring window, an automatic rollback is triggered. If the window expires without failure, the container is considered stable.

Lifecycle Hooks

You can run commands before and after a container update using container labels:

Hooks execute via /bin/sh -c inside the Drydock container. The following environment variables are available to hook commands:

A non-zero exit code or timeout is treated as failure. When dd.hook.pre.abort is true (the default), a failed pre-hook prevents the update from proceeding.

Self-Update

Drydock can update itself. When it detects that its own container has an available update, it performs a safe self-replacement sequence:

1. Notifies connected UI clients (full-screen overlay)
2. Backs up the current image metadata
3. Pulls the new image while the current container is still running
4. Renames the old container to drydock-old-{timestamp}
5. Creates a new container with the original name and new image
6. Spawns a temporary helper container (with the Docker socket) that stops the old container and starts the new one
7. The UI auto-reconnects to the new instance

If any step fails, Drydock falls back to the previous container automatically.

Examples

services:
  drydock:
    image: codeswhat/drydock
    ...
    environment:
      - DD_TRIGGER_DOCKER_EXAMPLE_PRUNE=true

docker run \
  -e "DD_TRIGGER_DOCKER_EXAMPLE_PRUNE=true" \
  ...
  codeswhat/drydock

Auto-rollback with health checks

Add these labels to a container to enable automatic rollback if it becomes unhealthy after an update:

services:
  myapp:
    image: myapp:latest
    labels:
      - dd.watch=true
      - dd.rollback.auto=true
      - dd.rollback.window=300000
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8080/health"]
      interval: 10s
      timeout: 5s
      retries: 3

Lifecycle hooks

Run a database migration before updating and a cache clear after:

services:
  myapp:
    image: myapp:latest
    labels:
      - dd.watch=true
      - dd.hook.pre=curl -X POST http://myapp:8080/prepare-update
      - dd.hook.post=curl -X POST http://myapp:8080/clear-cache
      - dd.hook.pre.abort=true
      - dd.hook.timeout=30000