Merge pull request 'Add --update flag to deploy create' (#984) from roysc/deployment-create-sync into main

Reviewed-on: #984
Reviewed-by: AFDudley <afdudley@noreply.git.vdb.to>

docs/cli.md

@@ -65,3 +65,71 @@ Force full rebuild of packages:
 ```
 $ 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 `deploy up` and `deploy down`.
+
+### deploy init
+
+Generate a deployment spec file from a stack definition:
+```
+$ laconic-so --stack <stack-name> deploy init --output <spec-file>
+```
+
+Options:
+- `--output` (required): write spec file here
+- `--config`: provide config variables for the deployment
+- `--config-file`: provide config variables in a file
+- `--kube-config`: provide a config file for a k8s deployment
+- `--image-registry`: provide a container image registry url for this k8s cluster
+- `--map-ports-to-host`: map ports to the host (`any-variable-random`, `localhost-same`, `any-same`, `localhost-fixed-random`, `any-fixed-random`)
+
+### deploy create
+
+Create a deployment directory from a spec file:
+```
+$ laconic-so --stack <stack-name> deploy create --spec-file <spec-file> --deployment-dir <dir>
+```
+
+Update an existing deployment in-place (preserving data volumes and env file):
+```
+$ laconic-so --stack <stack-name> deploy create --spec-file <spec-file> --deployment-dir <dir> --update
+```
+
+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.
+- `--network-dir`: network configuration supplied in this directory
+- `--initial-peers`: initial set of persistent peers
+
+### deploy up
+
+Start a deployment:
+```
+$ laconic-so deployment --dir <deployment-dir> up
+```
+
+### deploy down
+
+Stop a deployment:
+```
+$ laconic-so deployment --dir <deployment-dir> down
+```
+Use `--delete-volumes` to also remove data volumes.
+
+### deploy ps
+
+Show running services:
+```
+$ laconic-so deployment --dir <deployment-dir> ps
+```
+
+### deploy logs
+
+View service logs:
+```
+$ laconic-so deployment --dir <deployment-dir> logs
+```
+Use `-f` to follow and `-n <count>` to tail.

stack_orchestrator/data/compose/docker-compose-test.yml

@@ -8,6 +8,8 @@ services:
       CERC_TEST_PARAM_2: "CERC_TEST_PARAM_2_VALUE"
       CERC_TEST_PARAM_3: ${CERC_TEST_PARAM_3:-FAILED}
     volumes:
+      - ../config/test/script.sh:/opt/run.sh
+      - ../config/test/settings.env:/opt/settings.env
       - test-data-bind:/data
       - test-data-auto:/data2
       - test-config:/config:ro

stack_orchestrator/data/config/test/script.sh

@@ -0,0 +1,3 @@
+#!/bin/sh
+
+echo "Hello"

stack_orchestrator/data/config/test/settings.env

@@ -0,0 +1 @@
+ANSWER=42

stack_orchestrator/data/container-build/cerc-test-container/Dockerfile

@@ -1,9 +1,6 @@
-FROM ubuntu:latest
+FROM alpine:latest
 
-RUN apt-get update && export DEBIAN_FRONTEND=noninteractive && export DEBCONF_NOWARNINGS="yes" && \
-    apt-get install -y software-properties-common && \
-    apt-get install -y nginx && \
-    apt-get clean && rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/*
+RUN apk add --no-cache nginx
 
 EXPOSE 80
 

stack_orchestrator/data/container-build/cerc-test-container/run.sh

@@ -1,4 +1,4 @@
-#!/usr/bin/env bash
+#!/usr/bin/env sh
 set -e
 
 if [ -n "$CERC_SCRIPT_DEBUG" ]; then
@@ -8,14 +8,14 @@ fi
 echo "Test container starting"
 
 DATA_DEVICE=$(df | grep "/data$" | awk '{ print $1 }')
-if [[ -n "$DATA_DEVICE" ]]; then
+if [ -n "$DATA_DEVICE" ]; then
   echo "/data: MOUNTED dev=${DATA_DEVICE}"
 else
   echo "/data: not mounted"
 fi
 
 DATA2_DEVICE=$(df | grep "/data2$" | awk '{ print $1 }')
-if [[ -n "$DATA_DEVICE" ]]; then
+if [ -n "$DATA_DEVICE" ]; then
   echo "/data2: MOUNTED dev=${DATA2_DEVICE}"
 else
   echo "/data2: not mounted"
@@ -23,7 +23,7 @@ fi
 
 # Test if the container's filesystem is old (run previously) or new
 for d in /data /data2; do
-  if [[ -f "$d/exists" ]];
+  if [ -f "$d/exists" ];
   then
       TIMESTAMP=`cat $d/exists`
       echo "$d filesystem is old, created: $TIMESTAMP"
@@ -52,7 +52,7 @@ fi
 if [ -d "/config" ]; then
   echo "/config: EXISTS"
   for f in /config/*; do
-    if [[ -f "$f" ]] || [[ -L "$f" ]]; then
+    if [ -f "$f" ] || [ -L "$f" ]; then
       echo "$f:"
       cat "$f"
       echo ""
@@ -64,4 +64,4 @@ else
 fi
 
 # Run nginx which will block here forever
-/usr/sbin/nginx -g "daemon off;"
+nginx -g "daemon off;"

stack_orchestrator/deploy/deployment_create.py

@@ -19,9 +19,12 @@ import os
 from pathlib import Path
 from typing import List
 import random
-from shutil import copy, copyfile, copytree
+from shutil import copy, copyfile, copytree, rmtree
 from secrets import token_hex
 import sys
+import filecmp
+import tempfile
+
 from stack_orchestrator import constants
 from stack_orchestrator.opts import opts
 from stack_orchestrator.util import (
@@ -524,6 +527,12 @@ def _check_volume_definitions(spec):
     "--spec-file", required=True, help="Spec file to use to create this deployment"
 )
 @click.option("--deployment-dir", help="Create deployment files in this directory")
+@click.option(
+    "--update",
+    is_flag=True,
+    default=False,
+    help="Update existing deployment directory, preserving data volumes and env file",
+)
 @click.option(
     "--helm-chart",
     is_flag=True,
@@ -536,13 +545,21 @@ def _check_volume_definitions(spec):
 @click.argument("extra_args", nargs=-1, type=click.UNPROCESSED)
 @click.pass_context
 def create(
-    ctx, spec_file, deployment_dir, helm_chart, network_dir, initial_peers, extra_args
+    ctx,
+    spec_file,
+    deployment_dir,
+    update,
+    helm_chart,
+    network_dir,
+    initial_peers,
+    extra_args,
 ):
     deployment_command_context = ctx.obj
     return create_operation(
         deployment_command_context,
         spec_file,
         deployment_dir,
+        update,
         helm_chart,
         network_dir,
         initial_peers,
@@ -556,6 +573,7 @@ def create_operation(
     deployment_command_context,
     spec_file,
     deployment_dir,
+    update=False,
     helm_chart=False,
     network_dir=None,
     initial_peers=None,
@@ -568,23 +586,23 @@ def create_operation(
     stack_name = parsed_spec["stack"]
     deployment_type = parsed_spec[constants.deploy_to_key]
 
-    stack_file = get_stack_path(stack_name).joinpath(constants.stack_file_name)
-    parsed_stack = get_parsed_stack_config(stack_name)
     if opts.o.debug:
         print(f"parsed spec: {parsed_spec}")
+
     if deployment_dir is None:
         deployment_dir_path = _make_default_deployment_dir()
     else:
         deployment_dir_path = Path(deployment_dir)
-    if deployment_dir_path.exists():
-        error_exit(f"{deployment_dir_path} already exists")
-    os.mkdir(deployment_dir_path)
-    # Copy spec file and the stack file into the deployment dir
-    copyfile(spec_file, deployment_dir_path.joinpath(constants.spec_file_name))
-    copyfile(stack_file, deployment_dir_path.joinpath(constants.stack_file_name))
 
-    # Create deployment.yml with cluster-id
-    _create_deployment_file(deployment_dir_path)
+    if deployment_dir_path.exists():
+        if not update:
+            error_exit(f"{deployment_dir_path} already exists")
+        if opts.o.debug:
+            print(f"Updating existing deployment at {deployment_dir_path}")
+    else:
+        if update:
+            error_exit(f"--update requires that {deployment_dir_path} already exists")
+        os.mkdir(deployment_dir_path)
 
     # Branch to Helm chart generation flow if --helm-chart flag is set
     if deployment_type == "k8s" and helm_chart:
@@ -595,104 +613,41 @@ def create_operation(
         generate_helm_chart(stack_name, spec_file, deployment_dir_path)
         return  # Exit early for helm chart generation
 
-    # Existing deployment flow continues unchanged
-    # Copy any config varibles from the spec file into an env file suitable for compose
-    _write_config_file(
-        spec_file, deployment_dir_path.joinpath(constants.config_file_name)
-    )
-    # Copy any k8s config file into the deployment dir
-    if deployment_type == "k8s":
-        _write_kube_config_file(
-            Path(parsed_spec[constants.kube_config_key]),
-            deployment_dir_path.joinpath(constants.kube_config_filename),
-        )
-    # Copy the pod files into the deployment dir, fixing up content
-    pods = get_pod_list(parsed_stack)
-    destination_compose_dir = deployment_dir_path.joinpath("compose")
-    os.mkdir(destination_compose_dir)
-    destination_pods_dir = deployment_dir_path.joinpath("pods")
-    os.mkdir(destination_pods_dir)
-    yaml = get_yaml()
-    for pod in pods:
-        pod_file_path = get_pod_file_path(stack_name, parsed_stack, pod)
-        if pod_file_path is None:
-            continue
-        parsed_pod_file = yaml.load(open(pod_file_path, "r"))
-        extra_config_dirs = _find_extra_config_dirs(parsed_pod_file, pod)
-        destination_pod_dir = destination_pods_dir.joinpath(pod)
-        os.mkdir(destination_pod_dir)
-        if opts.o.debug:
-            print(f"extra config dirs: {extra_config_dirs}")
-        _fixup_pod_file(parsed_pod_file, parsed_spec, destination_compose_dir)
-        with open(
-            destination_compose_dir.joinpath("docker-compose-%s.yml" % pod), "w"
-        ) as output_file:
-            yaml.dump(parsed_pod_file, output_file)
-        # Copy the config files for the pod, if any
-        config_dirs = {pod}
-        config_dirs = config_dirs.union(extra_config_dirs)
-        for config_dir in config_dirs:
-            source_config_dir = resolve_config_dir(stack_name, config_dir)
-            if os.path.exists(source_config_dir):
-                destination_config_dir = deployment_dir_path.joinpath(
-                    "config", config_dir
-                )
-                # If the same config dir appears in multiple pods, it may already have
-                # been copied
-                if not os.path.exists(destination_config_dir):
-                    copytree(source_config_dir, destination_config_dir)
-        # Copy the script files for the pod, if any
-        if pod_has_scripts(parsed_stack, pod):
-            destination_script_dir = destination_pod_dir.joinpath("scripts")
-            os.mkdir(destination_script_dir)
-            script_paths = get_pod_script_paths(parsed_stack, pod)
-            _copy_files_to_directory(script_paths, destination_script_dir)
-        if parsed_spec.is_kubernetes_deployment():
-            for configmap in parsed_spec.get_configmaps():
-                source_config_dir = resolve_config_dir(stack_name, configmap)
-                if os.path.exists(source_config_dir):
-                    destination_config_dir = deployment_dir_path.joinpath(
-                        "configmaps", configmap
-                    )
-                    copytree(
-                        source_config_dir, destination_config_dir, dirs_exist_ok=True
-                    )
-        else:
-            # TODO: We should probably only do this if the volume is marked :ro.
-            for volume_name, volume_path in parsed_spec.get_volumes().items():
-                source_config_dir = resolve_config_dir(stack_name, volume_name)
-                # Only copy if the source exists and is _not_ empty.
-                if os.path.exists(source_config_dir) and os.listdir(source_config_dir):
-                    destination_config_dir = deployment_dir_path.joinpath(volume_path)
-                    # Only copy if the destination exists and _is_ empty.
-                    if os.path.exists(destination_config_dir) and not os.listdir(
-                        destination_config_dir
-                    ):
-                        copytree(
-                            source_config_dir,
-                            destination_config_dir,
-                            dirs_exist_ok=True,
-                        )
+    if update:
+        # Sync mode: write to temp dir, then copy to deployment dir with backups
+        temp_dir = Path(tempfile.mkdtemp(prefix="deployment-sync-"))
+        try:
+            # Write deployment files to temp dir (skip deployment.yml to preserve cluster ID)
+            _write_deployment_files(
+                temp_dir,
+                Path(spec_file),
+                parsed_spec,
+                stack_name,
+                deployment_type,
+                include_deployment_file=False,
+            )
 
-    # Copy the job files into the deployment dir (for Docker deployments)
-    jobs = get_job_list(parsed_stack)
-    if jobs and not parsed_spec.is_kubernetes_deployment():
-        destination_compose_jobs_dir = deployment_dir_path.joinpath("compose-jobs")
-        os.mkdir(destination_compose_jobs_dir)
-        for job in jobs:
-            job_file_path = get_job_file_path(stack_name, parsed_stack, job)
-            if job_file_path and job_file_path.exists():
-                parsed_job_file = yaml.load(open(job_file_path, "r"))
-                _fixup_pod_file(parsed_job_file, parsed_spec, destination_compose_dir)
-                with open(
-                    destination_compose_jobs_dir.joinpath(
-                        "docker-compose-%s.yml" % job
-                    ),
-                    "w",
-                ) as output_file:
-                    yaml.dump(parsed_job_file, output_file)
-                if opts.o.debug:
-                    print(f"Copied job compose file: {job}")
+            # Copy from temp to deployment dir, excluding data volumes and backing up changed files
+            # Exclude data/* to avoid touching user data volumes
+            # Exclude config file to preserve deployment settings (XXX breaks passing config vars
+            # from spec. could warn about this or not exclude...)
+            exclude_patterns = ["data", "data/*", constants.config_file_name]
+            _safe_copy_tree(
+                temp_dir, deployment_dir_path, exclude_patterns=exclude_patterns
+            )
+        finally:
+            # Clean up temp dir
+            rmtree(temp_dir)
+    else:
+        # Normal mode: write directly to deployment dir
+        _write_deployment_files(
+            deployment_dir_path,
+            Path(spec_file),
+            parsed_spec,
+            stack_name,
+            deployment_type,
+            include_deployment_file=True,
+        )
 
     # Delegate to the stack's Python code
     # The deploy create command doesn't require a --stack argument so we need
@@ -712,6 +667,181 @@ def create_operation(
     )
 
 
+def _safe_copy_tree(src: Path, dst: Path, exclude_patterns: List[str] = None):
+    """
+    Recursively copy a directory tree, backing up changed files with .bak suffix.
+
+    :param src: Source directory
+    :param dst: Destination directory
+    :param exclude_patterns: List of path patterns to exclude (relative to src)
+    """
+    if exclude_patterns is None:
+        exclude_patterns = []
+
+    def should_exclude(path: Path) -> bool:
+        """Check if path matches any exclude pattern."""
+        rel_path = path.relative_to(src)
+        for pattern in exclude_patterns:
+            if rel_path.match(pattern):
+                return True
+        return False
+
+    def safe_copy_file(src_file: Path, dst_file: Path):
+        """Copy file, backing up destination if it differs."""
+        if (
+            dst_file.exists()
+            and not dst_file.is_dir()
+            and not filecmp.cmp(src_file, dst_file)
+        ):
+            os.rename(dst_file, f"{dst_file}.bak")
+        copy(src_file, dst_file)
+
+    # Walk the source tree
+    for src_path in src.rglob("*"):
+        if should_exclude(src_path):
+            continue
+
+        rel_path = src_path.relative_to(src)
+        dst_path = dst / rel_path
+
+        if src_path.is_dir():
+            dst_path.mkdir(parents=True, exist_ok=True)
+        else:
+            dst_path.parent.mkdir(parents=True, exist_ok=True)
+            safe_copy_file(src_path, dst_path)
+
+
+def _write_deployment_files(
+    target_dir: Path,
+    spec_file: Path,
+    parsed_spec: Spec,
+    stack_name: str,
+    deployment_type: str,
+    include_deployment_file: bool = True,
+):
+    """
+    Write deployment files to target directory.
+
+    :param target_dir: Directory to write files to
+    :param spec_file: Path to spec file
+    :param parsed_spec: Parsed spec object
+    :param stack_name: Name of stack
+    :param deployment_type: Type of deployment
+    :param include_deployment_file: Whether to create deployment.yml file (skip for update)
+    """
+    stack_file = get_stack_path(stack_name).joinpath(constants.stack_file_name)
+    parsed_stack = get_parsed_stack_config(stack_name)
+
+    # Copy spec file and the stack file into the target dir
+    copyfile(spec_file, target_dir.joinpath(constants.spec_file_name))
+    copyfile(stack_file, target_dir.joinpath(constants.stack_file_name))
+
+    # Create deployment file if requested
+    if include_deployment_file:
+        _create_deployment_file(target_dir)
+
+    # Copy any config variables from the spec file into an env file suitable for compose
+    _write_config_file(spec_file, target_dir.joinpath(constants.config_file_name))
+
+    # Copy any k8s config file into the target dir
+    if deployment_type == "k8s":
+        _write_kube_config_file(
+            Path(parsed_spec[constants.kube_config_key]),
+            target_dir.joinpath(constants.kube_config_filename),
+        )
+
+    # Copy the pod files into the target dir, fixing up content
+    pods = get_pod_list(parsed_stack)
+    destination_compose_dir = target_dir.joinpath("compose")
+    os.makedirs(destination_compose_dir, exist_ok=True)
+    destination_pods_dir = target_dir.joinpath("pods")
+    os.makedirs(destination_pods_dir, exist_ok=True)
+    yaml = get_yaml()
+
+    for pod in pods:
+        pod_file_path = get_pod_file_path(stack_name, parsed_stack, pod)
+        if pod_file_path is None:
+            continue
+        parsed_pod_file = yaml.load(open(pod_file_path, "r"))
+        extra_config_dirs = _find_extra_config_dirs(parsed_pod_file, pod)
+        destination_pod_dir = destination_pods_dir.joinpath(pod)
+        os.makedirs(destination_pod_dir, exist_ok=True)
+        if opts.o.debug:
+            print(f"extra config dirs: {extra_config_dirs}")
+        _fixup_pod_file(parsed_pod_file, parsed_spec, destination_compose_dir)
+        with open(
+            destination_compose_dir.joinpath("docker-compose-%s.yml" % pod), "w"
+        ) as output_file:
+            yaml.dump(parsed_pod_file, output_file)
+
+        # Copy the config files for the pod, if any
+        config_dirs = {pod}
+        config_dirs = config_dirs.union(extra_config_dirs)
+        for config_dir in config_dirs:
+            source_config_dir = resolve_config_dir(stack_name, config_dir)
+            if os.path.exists(source_config_dir):
+                destination_config_dir = target_dir.joinpath("config", config_dir)
+                copytree(source_config_dir, destination_config_dir, dirs_exist_ok=True)
+
+        # Copy the script files for the pod, if any
+        if pod_has_scripts(parsed_stack, pod):
+            destination_script_dir = destination_pod_dir.joinpath("scripts")
+            os.makedirs(destination_script_dir, exist_ok=True)
+            script_paths = get_pod_script_paths(parsed_stack, pod)
+            _copy_files_to_directory(script_paths, destination_script_dir)
+
+        if parsed_spec.is_kubernetes_deployment():
+            for configmap in parsed_spec.get_configmaps():
+                source_config_dir = resolve_config_dir(stack_name, configmap)
+                if os.path.exists(source_config_dir):
+                    destination_config_dir = target_dir.joinpath(
+                        "configmaps", configmap
+                    )
+                    copytree(
+                        source_config_dir, destination_config_dir, dirs_exist_ok=True
+                    )
+        else:
+            # TODO:
+            # this is odd - looks up config dir that matches a volume name, then copies as a mount dir?
+            # AFAICT this is not used by or relevant to any existing stack - roy
+
+            # TODO: We should probably only do this if the volume is marked :ro.
+            for volume_name, volume_path in parsed_spec.get_volumes().items():
+                source_config_dir = resolve_config_dir(stack_name, volume_name)
+                # Only copy if the source exists and is _not_ empty.
+                if os.path.exists(source_config_dir) and os.listdir(source_config_dir):
+                    destination_config_dir = target_dir.joinpath(volume_path)
+                    # Only copy if the destination exists and _is_ empty.
+                    if os.path.exists(destination_config_dir) and not os.listdir(
+                        destination_config_dir
+                    ):
+                        copytree(
+                            source_config_dir,
+                            destination_config_dir,
+                            dirs_exist_ok=True,
+                        )
+
+    # Copy the job files into the target dir (for Docker deployments)
+    jobs = get_job_list(parsed_stack)
+    if jobs and not parsed_spec.is_kubernetes_deployment():
+        destination_compose_jobs_dir = target_dir.joinpath("compose-jobs")
+        os.makedirs(destination_compose_jobs_dir, exist_ok=True)
+        for job in jobs:
+            job_file_path = get_job_file_path(stack_name, parsed_stack, job)
+            if job_file_path and job_file_path.exists():
+                parsed_job_file = yaml.load(open(job_file_path, "r"))
+                _fixup_pod_file(parsed_job_file, parsed_spec, destination_compose_dir)
+                with open(
+                    destination_compose_jobs_dir.joinpath(
+                        "docker-compose-%s.yml" % job
+                    ),
+                    "w",
+                ) as output_file:
+                    yaml.dump(parsed_job_file, output_file)
+                if opts.o.debug:
+                    print(f"Copied job compose file: {job}")
+
+
 # TODO: this code should be in the stack .py files but
 # we haven't yet figured out how to integrate click across
 # the plugin boundary

stack_orchestrator/deploy/k8s/helpers.py

@@ -27,6 +27,48 @@ from stack_orchestrator.deploy.deployer import DeployerException
 from stack_orchestrator import constants
 
 
+def is_host_path_mount(volume_name: str) -> bool:
+    """Check if a volume name is a host path mount (starts with /, ., or ~)."""
+    return volume_name.startswith(("/", ".", "~"))
+
+
+def sanitize_host_path_to_volume_name(host_path: str) -> str:
+    """Convert a host path to a valid k8s volume name.
+
+    K8s volume names must be lowercase, alphanumeric, with - allowed.
+    E.g., '../config/test/script.sh' -> 'host-path-config-test-script-sh'
+    """
+    # Remove leading ./ or ../
+    clean_path = re.sub(r"^\.+/", "", host_path)
+    # Replace path separators and dots with hyphens
+    name = re.sub(r"[/.]", "-", clean_path)
+    # Remove any non-alphanumeric characters except hyphens
+    name = re.sub(r"[^a-zA-Z0-9-]", "", name)
+    # Convert to lowercase
+    name = name.lower()
+    # Remove leading/trailing hyphens and collapse multiple hyphens
+    name = re.sub(r"-+", "-", name).strip("-")
+    # Prefix with 'host-path-' to distinguish from named volumes
+    return f"host-path-{name}"
+
+
+def resolve_host_path_for_kind(host_path: str, deployment_dir: Path) -> Path:
+    """Resolve a host path mount (relative to compose file) to absolute path.
+
+    Compose files are in deployment_dir/compose/, so '../config/foo'
+    resolves to deployment_dir/config/foo.
+    """
+    # The path is relative to the compose directory
+    compose_dir = deployment_dir.joinpath("compose")
+    resolved = compose_dir.joinpath(host_path).resolve()
+    return resolved
+
+
+def get_kind_host_path_mount_path(sanitized_name: str) -> str:
+    """Get the path inside the kind node where a host path mount will be available."""
+    return f"/mnt/{sanitized_name}"
+
+
 def get_kind_cluster():
     """Get an existing kind cluster, if any.
 
@@ -177,6 +219,7 @@ def volume_mounts_for_service(parsed_pod_files, service):
                         for mount_string in volumes:
                             # Looks like: test-data:/data
                             # or test-data:/data:ro or test-data:/data:rw
+                            # or ../config/file.sh:/opt/file.sh (host path mount)
                             if opts.o.debug:
                                 print(f"mount_string: {mount_string}")
                             mount_split = mount_string.split(":")
@@ -185,13 +228,21 @@ def volume_mounts_for_service(parsed_pod_files, service):
                             mount_options = (
                                 mount_split[2] if len(mount_split) == 3 else None
                             )
+                            # For host path mounts, use sanitized name
+                            if is_host_path_mount(volume_name):
+                                k8s_volume_name = sanitize_host_path_to_volume_name(
+                                    volume_name
+                                )
+                            else:
+                                k8s_volume_name = volume_name
                             if opts.o.debug:
                                 print(f"volume_name: {volume_name}")
+                                print(f"k8s_volume_name: {k8s_volume_name}")
                                 print(f"mount path: {mount_path}")
                                 print(f"mount options: {mount_options}")
                             volume_device = client.V1VolumeMount(
                                 mount_path=mount_path,
-                                name=volume_name,
+                                name=k8s_volume_name,
                                 read_only="ro" == mount_options,
                             )
                             result.append(volume_device)
@@ -200,8 +251,12 @@ def volume_mounts_for_service(parsed_pod_files, service):
 
 def volumes_for_pod_files(parsed_pod_files, spec, app_name):
     result = []
+    seen_host_path_volumes = set()  # Track host path volumes to avoid duplicates
+
     for pod in parsed_pod_files:
         parsed_pod_file = parsed_pod_files[pod]
+
+        # Handle named volumes from top-level volumes section
         if "volumes" in parsed_pod_file:
             volumes = parsed_pod_file["volumes"]
             for volume_name in volumes.keys():
@@ -220,6 +275,35 @@ def volumes_for_pod_files(parsed_pod_files, spec, app_name):
                         name=volume_name, persistent_volume_claim=claim
                     )
                     result.append(volume)
+
+        # Handle host path mounts from service volumes
+        if "services" in parsed_pod_file:
+            services = parsed_pod_file["services"]
+            for service_name in services:
+                service_obj = services[service_name]
+                if "volumes" in service_obj:
+                    for mount_string in service_obj["volumes"]:
+                        mount_split = mount_string.split(":")
+                        volume_source = mount_split[0]
+                        if is_host_path_mount(volume_source):
+                            sanitized_name = sanitize_host_path_to_volume_name(
+                                volume_source
+                            )
+                            if sanitized_name not in seen_host_path_volumes:
+                                seen_host_path_volumes.add(sanitized_name)
+                                # Create hostPath volume for mount inside kind node
+                                kind_mount_path = get_kind_host_path_mount_path(
+                                    sanitized_name
+                                )
+                                host_path_source = client.V1HostPathVolumeSource(
+                                    path=kind_mount_path, type="FileOrCreate"
+                                )
+                                volume = client.V1Volume(
+                                    name=sanitized_name, host_path=host_path_source
+                                )
+                                result.append(volume)
+                                if opts.o.debug:
+                                    print(f"Created hostPath volume: {sanitized_name}")
     return result
 
 
@@ -238,6 +322,8 @@ def _make_absolute_host_path(data_mount_path: Path, deployment_dir: Path) -> Pat
 def _generate_kind_mounts(parsed_pod_files, deployment_dir, deployment_context):
     volume_definitions = []
     volume_host_path_map = _get_host_paths_for_volumes(deployment_context)
+    seen_host_path_mounts = set()  # Track to avoid duplicate mounts
+
     # Note these paths are relative to the location of the pod files (at present)
     # So we need to fix up to make them correct and absolute because kind assumes
     # relative to the cwd.
@@ -252,28 +338,58 @@ def _generate_kind_mounts(parsed_pod_files, deployment_dir, deployment_context):
                     for mount_string in volumes:
                         # Looks like: test-data:/data
                         # or test-data:/data:ro or test-data:/data:rw
+                        # or ../config/file.sh:/opt/file.sh (host path mount)
                         if opts.o.debug:
                             print(f"mount_string: {mount_string}")
                         mount_split = mount_string.split(":")
                         volume_name = mount_split[0]
                         mount_path = mount_split[1]
-                        if opts.o.debug:
-                            print(f"volume_name: {volume_name}")
-                            print(f"map: {volume_host_path_map}")
-                            print(f"mount path: {mount_path}")
-                        if volume_name not in deployment_context.spec.get_configmaps():
-                            if volume_host_path_map[volume_name]:
-                                host_path = _make_absolute_host_path(
-                                    volume_host_path_map[volume_name],
-                                    deployment_dir,
+
+                        if is_host_path_mount(volume_name):
+                            # Host path mount - add extraMount for kind
+                            sanitized_name = sanitize_host_path_to_volume_name(
+                                volume_name
+                            )
+                            if sanitized_name not in seen_host_path_mounts:
+                                seen_host_path_mounts.add(sanitized_name)
+                                # Resolve path relative to compose directory
+                                host_path = resolve_host_path_for_kind(
+                                    volume_name, deployment_dir
                                 )
-                                container_path = get_kind_pv_bind_mount_path(
-                                    volume_name
+                                container_path = get_kind_host_path_mount_path(
+                                    sanitized_name
                                 )
                                 volume_definitions.append(
                                     f"  - hostPath: {host_path}\n"
                                     f"    containerPath: {container_path}\n"
                                 )
+                                if opts.o.debug:
+                                    print(f"Added host path mount: {host_path}")
+                        else:
+                            # Named volume
+                            if opts.o.debug:
+                                print(f"volume_name: {volume_name}")
+                                print(f"map: {volume_host_path_map}")
+                                print(f"mount path: {mount_path}")
+                            if (
+                                volume_name
+                                not in deployment_context.spec.get_configmaps()
+                            ):
+                                if (
+                                    volume_name in volume_host_path_map
+                                    and volume_host_path_map[volume_name]
+                                ):
+                                    host_path = _make_absolute_host_path(
+                                        volume_host_path_map[volume_name],
+                                        deployment_dir,
+                                    )
+                                    container_path = get_kind_pv_bind_mount_path(
+                                        volume_name
+                                    )
+                                    volume_definitions.append(
+                                        f"  - hostPath: {host_path}\n"
+                                        f"    containerPath: {container_path}\n"
+                                    )
     return (
         ""
         if len(volume_definitions) == 0

stack_orchestrator/deploy/webapp/deploy_webapp.py

@@ -94,7 +94,7 @@ def create_deployment(
     # Add the TLS and DNS spec
     _fixup_url_spec(spec_file_name, url)
     create_operation(
-        deploy_command_context, spec_file_name, deployment_dir, False, None, None
+        deploy_command_context, spec_file_name, deployment_dir, False, False, None, None
     )
     # Fix up the container tag inside the deployment compose file
     _fixup_container_tag(deployment_dir, image)

tests/database/run-test.sh

@@ -86,7 +86,7 @@ fi
 echo "deploy init test: passed"
 
 # Switch to a full path for the data dir so it gets provisioned as a host bind mounted volume and preserved beyond cluster lifetime
-sed -i "s|^\(\s*db-data:$\)$|\1 ${test_deployment_dir}/data/db-data|" $test_deployment_spec
+sed -i.bak "s|^\(\s*db-data:$\)$|\1 ${test_deployment_dir}/data/db-data|" $test_deployment_spec
 
 $TEST_TARGET_SO --stack ${stack} deploy create --spec-file $test_deployment_spec --deployment-dir $test_deployment_dir
 # Check the deployment dir exists

tests/deploy/run-deploy-test.sh

@@ -34,6 +34,7 @@ mkdir -p $CERC_REPO_BASE_DIR
 # with and without volume removal
 $TEST_TARGET_SO --stack test setup-repositories
 $TEST_TARGET_SO --stack test build-containers
+
 # Test deploy command execution
 $TEST_TARGET_SO --stack test deploy setup $CERC_REPO_BASE_DIR
 # Check that we now have the expected output directory
@@ -85,6 +86,7 @@ else
     exit 1
 fi
 $TEST_TARGET_SO --stack test deploy down --delete-volumes
+
 # Basic test of creating a deployment
 test_deployment_dir=$CERC_REPO_BASE_DIR/test-deployment-dir
 test_deployment_spec=$CERC_REPO_BASE_DIR/test-deployment-spec.yml
@@ -122,6 +124,101 @@ fi
 echo "dbfc7a4d-44a7-416d-b5f3-29842cc47650" > $test_deployment_dir/data/test-config/test_config
 
 echo "deploy create output file test: passed"
+
+# Test sync functionality: update deployment without destroying data
+# First, create a marker file in the data directory to verify it's preserved
+test_data_marker="$test_deployment_dir/data/test-data-bind/sync-test-marker.txt"
+echo "original-data-$(date +%s)" > "$test_data_marker"
+original_marker_content=$(<$test_data_marker)
+
+# Modify a config file in the deployment to differ from source (to test backup)
+test_config_file="$test_deployment_dir/config/test/settings.env"
+test_config_file_original_content=$(<$test_config_file)
+test_config_file_changed_content="ANSWER=69"
+echo "$test_config_file_changed_content" > "$test_config_file"
+
+# Check a config file that matches the source (to test no backup for unchanged files)
+test_unchanged_config="$test_deployment_dir/config/test/script.sh"
+
+# Modify spec file to simulate an update
+sed -i.bak 's/CERC_TEST_PARAM_3:/CERC_TEST_PARAM_3: FASTER/' $test_deployment_spec
+
+# Create/modify config.env to test it isn't overwritten during sync
+config_env_file="$test_deployment_dir/config.env"
+config_env_persistent_content="PERSISTENT_VALUE=should-not-be-overwritten-$(date +%s)"
+echo "$config_env_persistent_content" >> "$config_env_file"
+original_config_env_content=$(<$config_env_file)
+
+# Run sync to update deployment files without destroying data
+$TEST_TARGET_SO --stack test deploy create --spec-file $test_deployment_spec --deployment-dir $test_deployment_dir --update
+
+# Verify config.env was not overwritten
+synced_config_env_content=$(<$config_env_file)
+if [ "$synced_config_env_content" == "$original_config_env_content" ]; then
+    echo "deployment update test: config.env preserved - passed"
+else
+    echo "deployment update test: config.env was overwritten - FAILED"
+    echo "Expected: $original_config_env_content"
+    echo "Got: $synced_config_env_content"
+    exit 1
+fi
+
+# Verify the spec file was updated in deployment dir
+updated_deployed_spec=$(<$test_deployment_dir/spec.yml)
+if [[ "$updated_deployed_spec" == *"FASTER"* ]]; then
+    echo "deployment update test: spec file updated"
+else
+    echo "deployment update test: spec file not updated - FAILED"
+    exit 1
+fi
+
+# Verify changed config file was backed up
+test_config_backup="${test_config_file}.bak"
+if [ -f "$test_config_backup" ]; then
+    backup_content=$(<$test_config_backup)
+    if [ "$backup_content" == "$test_config_file_changed_content" ]; then
+        echo "deployment update test: changed config file backed up - passed"
+    else
+        echo "deployment update test: backup content incorrect - FAILED"
+        exit 1
+    fi
+else
+    echo "deployment update test: backup file not created for changed file - FAILED"
+    exit 1
+fi
+
+# Verify unchanged config file was NOT backed up
+test_unchanged_backup="$test_unchanged_config.bak"
+if [ -f "$test_unchanged_backup" ]; then
+    echo "deployment update test: backup created for unchanged file - FAILED"
+    exit 1
+else
+    echo "deployment update test: no backup for unchanged file - passed"
+fi
+
+# Verify the config file was updated from source
+updated_config_content=$(<$test_config_file)
+if [ "$updated_config_content" == "$test_config_file_original_content" ]; then
+    echo "deployment update test: config file updated from source - passed"
+else
+    echo "deployment update test: config file not updated correctly - FAILED"
+    exit 1
+fi
+
+# Verify the data marker file still exists with original content
+if [ ! -f "$test_data_marker" ]; then
+    echo "deployment update test: data file deleted - FAILED"
+    exit 1
+fi
+synced_marker_content=$(<$test_data_marker)
+if [ "$synced_marker_content" == "$original_marker_content" ]; then
+    echo "deployment update test: data preserved - passed"
+else
+    echo "deployment update test: data corrupted - FAILED"
+    exit 1
+fi
+echo "deployment update test: passed"
+
 # Try to start the deployment
 $TEST_TARGET_SO deployment --dir $test_deployment_dir start
 # Check logs command works

tests/external-stack/run-test.sh

@@ -125,6 +125,49 @@ fi
 echo "dbfc7a4d-44a7-416d-b5f3-29842cc47650" > $test_deployment_dir/data/test-config/test_config
 
 echo "deploy create output file test: passed"
+
+# Test sync functionality: update deployment without destroying data
+# First, create a marker file in the data directory to verify it's preserved
+test_data_marker="$test_deployment_dir/data/test-data/sync-test-marker.txt"
+mkdir -p "$test_deployment_dir/data/test-data"
+echo "external-stack-data-$(date +%s)" > "$test_data_marker"
+original_marker_content=$(<$test_data_marker)
+# Verify deployment file exists and preserve its cluster ID
+original_cluster_id=$(grep "cluster-id:" "$test_deployment_dir/deployment.yml" 2>/dev/null || echo "")
+# Modify spec file to simulate an update
+sed -i.bak 's/CERC_TEST_PARAM_1=PASSED/CERC_TEST_PARAM_1=UPDATED/' $test_deployment_spec
+# Run sync to update deployment files without destroying data
+$TEST_TARGET_SO_STACK deploy create --spec-file $test_deployment_spec --deployment-dir $test_deployment_dir --update
+# Verify the spec file was updated in deployment dir
+updated_deployed_spec=$(<$test_deployment_dir/spec.yml)
+if [[ "$updated_deployed_spec" == *"UPDATED"* ]]; then
+    echo "deploy sync test: spec file updated"
+else
+    echo "deploy sync test: spec file not updated - FAILED"
+    exit 1
+fi
+# Verify the data marker file still exists with original content
+if [ ! -f "$test_data_marker" ]; then
+    echo "deploy sync test: data file deleted - FAILED"
+    exit 1
+fi
+synced_marker_content=$(<$test_data_marker)
+if [ "$synced_marker_content" == "$original_marker_content" ]; then
+    echo "deploy sync test: data preserved - passed"
+else
+    echo "deploy sync test: data corrupted - FAILED"
+    exit 1
+fi
+# Verify cluster ID was preserved (not regenerated)
+new_cluster_id=$(grep "cluster-id:" "$test_deployment_dir/deployment.yml" 2>/dev/null || echo "")
+if [ -n "$original_cluster_id" ] && [ "$original_cluster_id" == "$new_cluster_id" ]; then
+    echo "deploy sync test: cluster ID preserved - passed"
+else
+    echo "deploy sync test: cluster ID not preserved - FAILED"
+    exit 1
+fi
+echo "deploy sync test: passed"
+
 # Try to start the deployment
 $TEST_TARGET_SO deployment --dir $test_deployment_dir start
 # Check logs command works