DeployStack doesn't recreate containers when environment config changes

[boroughnexus-cto]

Issue

When updating a Stack's config.environment via UpdateStack API call, followed by DeployStack, the container is not recreated with the new environment variables.

Steps to Reproduce

1. Update stack config with new environment variable:

{"type":"UpdateStack","params":{"id":"<stack_id>","config":{"environment":"NEW_VAR=true"}}}

2. Deploy the stack:

{"type":"DeployStack","params":{"stack":"<stack_id>"}}

3. Check running container:

docker exec <container> env | grep NEW_VAR
# Returns nothing - variable not present

Expected Behavior

DeployStack should detect that environment variables have changed and recreate the container (docker-compose up -d --force-recreate or equivalent).

Actual Behavior

Container continues running with old environment. The only way to apply the change is to manually restart/recreate the container.

Workaround

Manually recreate container:

docker-compose down && docker-compose up -d

Environment

• Komodo Core: ghcr.io/moghtech/komodo-core:latest
• Stack type: Git-synced (TKNet-Docker-Stacks repo)
• API endpoint: /execute with DeployStack operation

Impact

Medium - Environment config changes require manual intervention instead of being applied automatically through the API/UI.

Related

This may be related to how Komodo detects whether a stack needs redeployment. It might only check git hash changes, not config changes.

[madereddy]

Running into a similar issue requiring manually redeploying the stack.

Stack with config file defined.

Update config file on repo, repo clones, but stack does not redeploy.

Tried creating a procedure that checks if the stack has changed and changes to the config file do not trigger a redeploy, but a change to the compose file does trigger a redeploy.

[tylovejoy]

It would be nice if we could add --force-recreate to the redeploy.

[blackstormlab]

To the people here there is an option called destroy before redeploy it sounds like that is what you need.