Skip to main content
Learn how to deploy, update, and manage your infrastructure using kombify TechStack’s deployment pipeline.

​
Deployment workflow

​
Deploy a new stack

1

Create your spec file

Create or update your kombination.yaml:
kombination.yaml
stackkit: base-kit

nodes:
  - name: main-server
    type: local
    connection:
      host: 192.168.1.100
      user: root

services:
  traefik:
    enabled: true
  dokploy:
    enabled: true
2

Validate configuration

Validate your spec before deployment:
curl -X POST http://localhost:5260/api/v1/unifier/validate \
  -H "Content-Type: application/yaml" \
  --data-binary @kombination.yaml
Or via the web UI: Stack → Validate
3

Generate infrastructure code

Generate OpenTofu and Docker configurations:
curl -X POST http://localhost:5260/api/v1/unifier/generate \
  -H "Content-Type: application/yaml" \
  --data-binary @kombination.yaml
4

Preview changes

See what will be created before applying:
curl -X POST http://localhost:5260/api/v1/unifier/preview \
  -H "Content-Type: application/yaml" \
  --data-binary @kombination.yaml
5

Apply deployment

Deploy to your infrastructure:
curl -X POST http://localhost:5260/api/v1/stacks/{stack-id}/provision
This triggers the full deployment pipeline:
  1. Run pre-checks on target nodes
  2. Apply OpenTofu configuration
  3. Deploy Docker services
  4. Verify health checks

​
Deployment states

StateDescription
pendingDeployment queued
validatingValidating spec
generatingGenerating IaC
planningRunning tofu plan
applyingApplying changes
deployingDeploying services
verifyingRunning health checks
completedSuccessfully deployed
failedDeployment failed
rolled_backRolled back to previous state

​
Update an existing stack

Apply changes without downtime:
curl -X POST http://localhost:5260/api/v1/stacks/{stack-id}/deploy \
  -H "Content-Type: application/json" \
  -d '{"strategy": "rolling"}'

​
Destroy a stack

This will permanently destroy all resources. Make sure to backup any data first.
curl -X POST http://localhost:5260/api/v1/stacks/{stack-id}/destroy

​
Deployment jobs

All deployments run as background jobs. Check job status: 
# List recent jobs
curl http://localhost:5260/api/v1/jobs

# Get specific job status
curl http://localhost:5260/api/v1/jobs/{job-id}

​
Job status response

{
  "id": "job_abc123",
  "type": "provision",
  "status": "running",
  "progress": 65,
  "started_at": "2026-02-03T10:00:00Z",
  "steps": [
    {"name": "validate", "status": "completed"},
    {"name": "generate", "status": "completed"},
    {"name": "apply", "status": "running"},
    {"name": "verify", "status": "pending"}
  ]
}

​
Pre-deployment checks

Stack runs pre-checks before deployment to ensure nodes are ready:
CheckDescription
ConnectivitySSH/agent connection working
OS compatibilitySupported distribution and version
ResourcesSufficient CPU, RAM, disk
DockerDocker installed and running
PortsRequired ports available
# Run pre-checks manually
curl -X POST http://localhost:5260/api/v1/workers/{worker-id}/prechecks

​
Rollback

If a deployment fails, Stack can rollback to the previous state: 
# Automatic rollback (enabled by default)
curl -X POST http://localhost:5260/api/v1/stacks/{stack-id}/deploy \
  -d '{"auto_rollback": true}'

# Manual rollback
curl -X POST http://localhost:5260/api/v1/stacks/{stack-id}/rollback

​
Next steps

Monitoring

Monitor your deployed infrastructure

Drift detection

Detect and fix configuration drift