Blueprint · Go Container Worker

Go Container Worker Starter (work-go-minimal)

Minimal Go worker that responds with 'hello from worker' on every invoke. Run compiled Go binaries as managed container workers on KERNSTACK — built for heavy compute, streaming pipelines, and volume-mounted workloads.

Container Workers Object Buckets Edge Gateway

Talk to us Read Technical Documentation → Software Lifecycle Cloud →

One-Command CLI Initialization

CLI Command work-go-minimal

kernstack add worker --name processor --blueprint work-go-minimal

Built for workloads that outlive a single request

While WebAssembly functions handle rapid request-response edges, work-go-minimal gives your stack a dedicated Go container runtime for CPU-intensive pipelines, binary transformations, and persistent concurrency.

Workers live in the exact same stack.json as your functions, tables, and buckets, sharing the same identity model and single credit meter.

Native Go binary execution on dedicated European bare metal
Mount frozen bucket prefix snapshots as read-only local volumes
Invoke from functions, workflows, or gateway routes without VPC peering

Operational simplicity without Kubernetes or ECS

Product teams of two to eight engineers should not spend sprints maintaining ECS task definitions, Fargate networking, or Kubernetes ingress controllers.

With work-go-minimal, placement, health checking, log fanout, and scale-to-zero are handled by the KERNSTACK control plane.

When a Go worker is the right blueprint

work-go-minimal is a container worker for jobs that outlive one HTTP request: image transforms, archive builds, and consumers that walk a table. It keeps a filesystem volume when the job needs scratch space, and it shares the project tables and buckets with any WASM function in the same stack.json.

You do not schedule the container yourself. The worker entry names the image build and the route or queue that wakes it. kernstack plan shows whether the volume, the environment target, and the route changed. That is the review surface, instead of a Kubernetes deployment manifest plus a service object plus an autoscaler.

Starter Code — work-go-minimal go

package main

import (
  "context"
  "net/http"

  "github.com/coding-berlin/kernstack-sdk-go/sdk_workers/sdk_workers_runtime"
)

func handler(_ context.Context, req sdk_workers_runtime.WorkerInvokeRequest) (sdk_workers_runtime.WorkerInvokeResponse, error) {
  return sdk_workers_runtime.WorkerInvokeResponse{
    RequestID:  req.RequestID,
    StatusCode: http.StatusOK,
    Body:       []byte("hello from worker"),
    Headers:    map[string][]string{"Content-Type": {"text/plain"}},
  }, nil
}

func main() {
  sdk_workers_runtime.RunWithContext(context.Background(), handler)
}

stack.json — work-go-minimal plan · deploy

{
  "stack": { "target": "app.prod" },
  "resources": {
    "buckets": [
      {
        "name": "assets",
        "volumes": [{ "name": "models", "source_prefix": "weights/" }]
      }
    ],
    "workers": [
      {
        "name": "processor",
        "runtime": "binary",
        "volumes": [{ "name": "models" }]
      }
    ]
  }
}

FAQ

Frequently asked questions

How do I add a Go worker to my stack?

Run kernstack add worker --name processor --blueprint work-go-minimal. The CLI creates the worker directory, vendors the Go SDK, and registers the worker in stack.json.

Can Go workers mount large files from Buckets?

Yes. Declare a volume on any bucket prefix in stack.json and attach it to the worker to read large assets, datasets, or binaries from local disk.

How is worker execution metered?

Workers draw from your environment credit balance using the standard compute formula: RAM in GiB multiplied by wall-clock seconds at 0.01 credits per GiB-second.

Connected in the Project Graph

Related capabilities, architectures, and comparisons

Ship your stack on the Software Lifecycle Cloud.

Closed preview on dedicated European infrastructure. Bring a repository or start from a blueprint.

Talk to us See pricing & credits →