feat: add local test runner script

Creates a temporary venv, installs shiv, builds the laconic-so package,
and runs any test script in isolation. Cleans up on exit.

Usage: ./tests/scripts/run-test-local.sh tests/webapp-test/run-webapp-test.sh

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

TODO.md

@@ -7,6 +7,25 @@ We need an "update stack" command in stack orchestrator and cleaner documentatio
 
 **Context**: Currently, `deploy init` generates a spec file and `deploy create` creates a deployment directory. The `deployment update` command (added by Thomas Lackey) only syncs env vars and restarts - it doesn't regenerate configurations. There's a gap in the workflow for updating stack configurations after initial deployment.
 
+## Bugs
+
+### `deploy create` doesn't auto-generate volume mappings for new pods
+
+When a new pod is added to `stack.yml` (e.g. `monitoring`), `deploy create`
+does not generate default host path mappings in spec.yml for the new pod's
+volumes. The deployment then fails at scheduling because the PVCs don't exist.
+
+**Expected**: `deploy create` enumerates all volumes from all compose files
+in the stack and generates default host paths for any that aren't already
+mapped in the spec.yml `volumes:` section.
+
+**Actual**: Only volumes already in spec.yml get PVs. New volumes are silently
+missing, causing `FailedScheduling: persistentvolumeclaim not found`.
+
+**Workaround**: Manually add volume entries to spec.yml and create host dirs.
+
+**Files**: `deployment_create.py` (`_write_config_file`, volume handling)
+
 ## Architecture Refactoring
 
 ### Separate Deployer from Stack Orchestrator CLI

docs/cli.md

@@ -68,7 +68,7 @@ $ laconic-so build-npms --include <package-name> --force-rebuild
 
 ## deploy
 
-The `deploy` command group manages persistent deployments. The general workflow is `deploy init` to generate a spec file, then `deploy create` to create a deployment directory from the spec, then runtime commands like `deployment start` and `deployment stop`.
+The `deploy` command group manages persistent deployments. The general workflow is `deploy init` to generate a spec file, then `deploy create` to create a deployment directory from the spec, then runtime commands like `deploy up` and `deploy down`.
 
 ### deploy init
 
@@ -101,91 +101,35 @@ Options:
 - `--spec-file` (required): spec file to use
 - `--deployment-dir`: target directory for deployment files
 - `--update`: update an existing deployment directory, preserving data volumes and env file. Changed files are backed up with a `.bak` suffix. The deployment's `config.env` and `deployment.yml` are also preserved.
-- `--helm-chart`: generate Helm chart instead of deploying (k8s only)
 - `--network-dir`: network configuration supplied in this directory
 - `--initial-peers`: initial set of persistent peers
 
-## deployment
+### deploy up
 
-Runtime commands for managing a created deployment. Use `--dir` to specify the deployment directory.
-
-### deployment start
-
-Start a deployment (`up` is a legacy alias):
+Start a deployment:
 ```
-$ laconic-so deployment --dir <deployment-dir> start
+$ laconic-so deployment --dir <deployment-dir> up
 ```
 
-Options:
-- `--stay-attached` / `--detatch-terminal`: attach to container stdout (default: detach)
-- `--skip-cluster-management` / `--perform-cluster-management`: skip kind cluster creation/teardown (default: perform management). Only affects k8s-kind deployments. Use this when multiple stacks share a single cluster.
+### deploy down
 
-### deployment stop
-
-Stop a deployment (`down` is a legacy alias):
+Stop a deployment:
 ```
-$ laconic-so deployment --dir <deployment-dir> stop
+$ laconic-so deployment --dir <deployment-dir> down
 ```
+Use `--delete-volumes` to also remove data volumes.
 
-Options:
-- `--delete-volumes` / `--preserve-volumes`: delete data volumes on stop (default: preserve)
-- `--skip-cluster-management` / `--perform-cluster-management`: skip kind cluster teardown (default: perform management). Use this to stop a single deployment without destroying a shared cluster.
-
-### deployment restart
-
-Restart a deployment with GitOps-aware workflow. Pulls latest stack code, syncs the deployment directory from the git-tracked spec, and restarts services:
-```
-$ laconic-so deployment --dir <deployment-dir> restart
-```
-
-See [deployment_patterns.md](deployment_patterns.md) for the recommended GitOps workflow.
-
-### deployment ps
+### deploy ps
 
 Show running services:
 ```
 $ laconic-so deployment --dir <deployment-dir> ps
 ```
 
-### deployment logs
+### deploy logs
 
 View service logs:
 ```
 $ laconic-so deployment --dir <deployment-dir> logs
 ```
 Use `-f` to follow and `-n <count>` to tail.
-
-### deployment exec
-
-Execute a command in a running service container:
-```
-$ laconic-so deployment --dir <deployment-dir> exec <service-name> "<command>"
-```
-
-### deployment status
-
-Show deployment status:
-```
-$ laconic-so deployment --dir <deployment-dir> status
-```
-
-### deployment port
-
-Show mapped ports for a service:
-```
-$ laconic-so deployment --dir <deployment-dir> port <service-name> <port>
-```
-
-### deployment push-images
-
-Push deployment images to a registry:
-```
-$ laconic-so deployment --dir <deployment-dir> push-images
-```
-
-### deployment run-job
-
-Run a one-time job in the deployment:
-```
-$ laconic-so deployment --dir <deployment-dir> run-job <job-name>
-```

docs/deployment_patterns.md

@@ -30,7 +30,7 @@ git commit -m "Add my-stack deployment configuration"
 git push
 
 # On deployment server: deploy from git-tracked spec
-laconic-so --stack my-stack deploy create \
+laconic-so deploy create \
   --spec-file /path/to/operator-repo/spec.yml \
   --deployment-dir my-deployment
 

stack_orchestrator/constants.py

@@ -29,7 +29,6 @@ network_key = "network"
 http_proxy_key = "http-proxy"
 image_registry_key = "image-registry"
 configmaps_key = "configmaps"
-secrets_key = "secrets"
 resources_key = "resources"
 volumes_key = "volumes"
 security_key = "security"

stack_orchestrator/deploy/deployment_create.py

@@ -477,9 +477,6 @@ def init_operation(
             spec_file_content["volumes"] = {**volume_descriptors, **orig_volumes}
         if configmap_descriptors:
             spec_file_content["configmaps"] = configmap_descriptors
-        if "k8s" in deployer_type:
-            if "secrets" not in spec_file_content:
-                spec_file_content["secrets"] = {}
 
     if opts.o.debug:
         print(

stack_orchestrator/deploy/k8s/cluster_info.py

@@ -394,13 +394,43 @@ class ClusterInfo:
             result.append(pv)
         return result
 
+    def _any_service_has_host_network(self):
+        for pod_name in self.parsed_pod_yaml_map:
+            pod = self.parsed_pod_yaml_map[pod_name]
+            for svc in pod.get("services", {}).values():
+                if svc.get("network_mode") == "host":
+                    return True
+        return False
+
+    def _resolve_container_resources(
+        self, container_name: str, service_info: dict, global_resources: Resources
+    ) -> Resources:
+        """Resolve resources for a container using layered priority.
+
+        Priority: spec per-container > compose deploy.resources
+        > spec global > DEFAULT
+        """
+        # 1. Check spec.yml for per-container override
+        per_container = self.spec.get_container_resources_for(container_name)
+        if per_container:
+            return per_container
+
+        # 2. Check compose service_info for deploy.resources
+        deploy_block = service_info.get("deploy", {})
+        compose_resources = deploy_block.get("resources", {}) if deploy_block else {}
+        if compose_resources:
+            return Resources(compose_resources)
+
+        # 3. Fall back to spec.yml global (already resolved with DEFAULT fallback)
+        return global_resources
+
     # TODO: put things like image pull policy into an object-scope struct
     def get_deployment(self, image_pull_policy: Optional[str] = None):
         containers = []
         services = {}
-        resources = self.spec.get_container_resources()
-        if not resources:
-            resources = DEFAULT_CONTAINER_RESOURCES
+        global_resources = self.spec.get_container_resources()
+        if not global_resources:
+            global_resources = DEFAULT_CONTAINER_RESOURCES
         for pod_name in self.parsed_pod_yaml_map:
             pod = self.parsed_pod_yaml_map[pod_name]
             services = pod["services"]
@@ -483,16 +513,9 @@ class ClusterInfo:
                         )
                     )
                 ]
-                # Mount user-declared secrets from spec.yml
-                for user_secret_name in self.spec.get_secrets():
-                    env_from.append(
-                        client.V1EnvFromSource(
-                            secret_ref=client.V1SecretEnvSource(
-                                name=user_secret_name,
-                                optional=True,
-                            )
-                        )
-                    )
+                container_resources = self._resolve_container_resources(
+                    container_name, service_info, global_resources
+                )
                 container = client.V1Container(
                     name=container_name,
                     image=image_to_use,
@@ -511,7 +534,7 @@ class ClusterInfo:
                         if self.spec.get_capabilities()
                         else None,
                     ),
-                    resources=to_k8s_resource_requirements(resources),
+                    resources=to_k8s_resource_requirements(container_resources),
                 )
                 containers.append(container)
         volumes = volumes_for_pod_files(
@@ -578,6 +601,7 @@ class ClusterInfo:
                     )
                 )
 
+        use_host_network = self._any_service_has_host_network()
         template = client.V1PodTemplateSpec(
             metadata=client.V1ObjectMeta(annotations=annotations, labels=labels),
             spec=client.V1PodSpec(
@@ -587,6 +611,8 @@ class ClusterInfo:
                 affinity=affinity,
                 tolerations=tolerations,
                 runtime_class_name=self.spec.get_runtime_class(),
+                host_network=use_host_network or None,
+                dns_policy=("ClusterFirstWithHostNet" if use_host_network else None),
             ),
         )
         spec = client.V1DeploymentSpec(

stack_orchestrator/deploy/spec.py

@@ -115,14 +115,32 @@ class Spec:
     def get_configmaps(self):
         return self.obj.get(constants.configmaps_key, {})
 
-    def get_secrets(self):
-        return self.obj.get(constants.secrets_key, {})
-
     def get_container_resources(self):
         return Resources(
             self.obj.get(constants.resources_key, {}).get("containers", {})
         )
 
+    def get_container_resources_for(
+        self, container_name: str
+    ) -> typing.Optional[Resources]:
+        """Look up per-container resource overrides from spec.yml.
+
+        Checks resources.containers.<container_name> in the spec. Returns None
+        if no per-container override exists (caller falls back to other sources).
+        """
+        containers_block = self.obj.get(constants.resources_key, {}).get(
+            "containers", {}
+        )
+        if container_name in containers_block:
+            entry = containers_block[container_name]
+            # Only treat it as a per-container override if it's a dict with
+            # reservations/limits nested inside (not a top-level global key)
+            if isinstance(entry, dict) and (
+                "reservations" in entry or "limits" in entry
+            ):
+                return Resources(entry)
+        return None
+
     def get_volume_resources(self):
         return Resources(
             self.obj.get(constants.resources_key, {}).get(constants.volumes_key, {})

tests/scripts/run-test-local.sh

@@ -0,0 +1,53 @@
+#!/bin/bash
+# Run a test suite locally in an isolated venv.
+#
+# Usage:
+#   ./tests/scripts/run-test-local.sh <test-script>
+#
+# Examples:
+#   ./tests/scripts/run-test-local.sh tests/webapp-test/run-webapp-test.sh
+#   ./tests/scripts/run-test-local.sh tests/smoke-test/run-smoke-test.sh
+#   ./tests/scripts/run-test-local.sh tests/k8s-deploy/run-deploy-test.sh
+#
+# The script creates a temporary venv, installs shiv, builds the laconic-so
+# package, runs the requested test, then cleans up.
+
+set -euo pipefail
+
+if [ $# -lt 1 ]; then
+  echo "Usage: $0 <test-script> [args...]"
+  exit 1
+fi
+
+TEST_SCRIPT="$1"
+shift
+
+if [ ! -f "$TEST_SCRIPT" ]; then
+  echo "Error: $TEST_SCRIPT not found"
+  exit 1
+fi
+
+REPO_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)"
+VENV_DIR=$(mktemp -d /tmp/so-test-XXXXXX)
+
+cleanup() {
+  echo "Cleaning up venv: $VENV_DIR"
+  rm -rf "$VENV_DIR"
+}
+trap cleanup EXIT
+
+cd "$REPO_DIR"
+
+echo "==> Creating venv in $VENV_DIR"
+python3 -m venv "$VENV_DIR"
+source "$VENV_DIR/bin/activate"
+
+echo "==> Installing shiv"
+pip install -q shiv
+
+echo "==> Building laconic-so package"
+./scripts/create_build_tag_file.sh
+./scripts/build_shiv_package.sh
+
+echo "==> Running: $TEST_SCRIPT $*"
+exec "./$TEST_SCRIPT" "$@"