adding podman support in docker configuration (#999)

Signed-off-by: Paige Patton <prubenda@redhat.com>

.github/workflows/tests.yml

@@ -101,6 +101,7 @@ jobs:
             echo "test_io_hog" >> ./CI/tests/functional_tests
             echo "test_pod_network_filter" >> ./CI/tests/functional_tests
             echo "test_pod_server" >> ./CI/tests/functional_tests
+            echo "test_node" >> ./CI/tests/functional_tests
             echo "test_pvc" >> ./CI/tests/functional_tests
 
       # Push on main only steps + all other functional to collect coverage
@@ -138,6 +139,7 @@ jobs:
           echo "test_io_hog" >> ./CI/tests/functional_tests
           echo "test_pod_network_filter" >> ./CI/tests/functional_tests
           echo "test_pod_server" >> ./CI/tests/functional_tests
+          echo "test_node" >> ./CI/tests/functional_tests
           echo "test_pvc" >> ./CI/tests/functional_tests
       # Final common steps
       - name: Run Functional tests

CI/tests/test_node.sh

@@ -0,0 +1,18 @@
+uset -xeEo pipefail
+
+source CI/tests/common.sh
+
+trap error ERR
+trap finish EXIT
+
+function functional_test_node_stop_start {
+  export scenario_type="node_scenarios"
+  export scenario_file="scenarios/kind/node_scenarios_example.yml"
+  export post_config=""
+  envsubst < CI/config/common_test_config.yaml > CI/config/node_config.yaml
+  cat CI/config/node_config.yaml
+  python3 -m coverage run -a run_kraken.py -c CI/config/node_config.yaml
+  echo "Node Stop/Start scenario test: Success"
+}
+
+functional_test_node_stop_start

krkn/scenario_plugins/node_actions/docker_node_scenarios.py

@@ -2,46 +2,173 @@ import krkn.scenario_plugins.node_actions.common_node_functions as nodeaction
 from krkn.scenario_plugins.node_actions.abstract_node_scenarios import (
     abstract_node_scenarios,
 )
+import os
+import platform
 import logging
 import docker
 from krkn_lib.k8s import KrknKubernetes
 from krkn_lib.models.k8s import AffectedNode, AffectedNodeStatus
 
 class Docker:
+    """
+    Container runtime client wrapper supporting both Docker and Podman.
+
+    This class automatically detects and connects to either Docker or Podman
+    container runtimes using the Docker-compatible API. It tries multiple
+    connection methods in order of preference:
+
+    1. Docker Unix socket (unix:///var/run/docker.sock)
+    2. Platform-specific Podman sockets:
+       - macOS: ~/.local/share/containers/podman/machine/podman.sock
+       - Linux rootful: unix:///run/podman/podman.sock
+       - Linux rootless: unix:///run/user/<uid>/podman/podman.sock
+    3. Environment variables (DOCKER_HOST or CONTAINER_HOST)
+
+    The runtime type (docker/podman) is auto-detected and logged for debugging.
+    Supports Kind clusters running on Podman.
+
+    Assisted By: Claude Code
+    """
     def __init__(self):
-        self.client = docker.from_env()
+        self.client = None
+        self.runtime = 'unknown'
+        
+
+        # Try multiple connection methods in order of preference
+        # Supports both Docker and Podman
+        connection_methods = [
+            ('unix:///var/run/docker.sock', 'Docker Unix socket'),
+        ]
+
+        # Add platform-specific Podman sockets
+        if platform.system() == 'Darwin':  # macOS
+            # On macOS, Podman uses podman-machine with socket typically at:
+            # ~/.local/share/containers/podman/machine/podman.sock
+            # This is often symlinked to /var/run/docker.sock
+            podman_machine_sock = os.path.expanduser('~/.local/share/containers/podman/machine/podman.sock')
+            if os.path.exists(podman_machine_sock):
+                connection_methods.append((f'unix://{podman_machine_sock}', 'Podman machine socket (macOS)'))
+        else:  # Linux
+            connection_methods.extend([
+                ('unix:///run/podman/podman.sock', 'Podman Unix socket (rootful)'),
+                ('unix:///run/user/{uid}/podman/podman.sock', 'Podman Unix socket (rootless)'),
+            ])
+
+        # Always try from_env as last resort
+        connection_methods.append(('from_env', 'Environment variables (DOCKER_HOST/CONTAINER_HOST)'))
+
+        for method, description in connection_methods:
+            try:
+                # Handle rootless Podman socket path with {uid} placeholder
+                if '{uid}' in method:
+                    uid = os.getuid()
+                    method = method.format(uid=uid)
+                    logging.info(f'Attempting to connect using {description}: {method}')
+
+                if method == 'from_env':
+                    logging.info(f'Attempting to connect using {description}')
+                    self.client = docker.from_env()
+                else:
+                    logging.info(f'Attempting to connect using {description}: {method}')
+                    self.client = docker.DockerClient(base_url=method)
+
+                # Test the connection
+                self.client.ping()
+
+                # Detect runtime type
+                try:
+                    version_info = self.client.version()
+                    version_str = version_info.get('Version', '')
+                    if 'podman' in version_str.lower():
+                        self.runtime = 'podman'
+                    else:
+                        self.runtime = 'docker'
+                    logging.debug(f'Runtime version info: {version_str}')
+                except Exception as version_err:
+                    logging.warning(f'Could not detect runtime version: {version_err}')
+                    self.runtime = 'unknown'
+
+                logging.info(f'Successfully connected to {self.runtime} using {description}')
+
+                # Log available containers for debugging
+                try:
+                    containers = self.client.containers.list(all=True)
+                    logging.info(f'Found {len(containers)} total containers')
+                    for container in containers[:5]:  # Log first 5
+                        logging.debug(f'  Container: {container.name} ({container.status})')
+                except Exception as list_err:
+                    logging.warning(f'Could not list containers: {list_err}')
+
+                break
+
+            except Exception as e:
+                logging.warning(f'Failed to connect using {description}: {e}')
+                continue
+
+        if self.client is None:
+            error_msg = 'Failed to initialize container runtime client (Docker/Podman) with any connection method'
+            logging.error(error_msg)
+            logging.error('Attempted connection methods:')
+            for method, desc in connection_methods:
+                logging.error(f'  - {desc}: {method}')
+            raise RuntimeError(error_msg)
+
+        logging.info(f'Container runtime client initialized successfully: {self.runtime}')
 
     def get_container_id(self, node_name):
+        """Get the container ID for a given node name."""
         container = self.client.containers.get(node_name)
+        logging.info(f'Found {self.runtime} container for node {node_name}: {container.id}')
         return container.id
 
     # Start the node instance
     def start_instances(self, node_name):
+        """Start a container instance (works with both Docker and Podman)."""
+        logging.info(f'Starting {self.runtime} container for node: {node_name}')
         container = self.client.containers.get(node_name)
         container.start()
+        logging.info(f'Container {container.id} started successfully')
 
     # Stop the node instance
     def stop_instances(self, node_name):
+        """Stop a container instance (works with both Docker and Podman)."""
+        logging.info(f'Stopping {self.runtime} container for node: {node_name}')
         container = self.client.containers.get(node_name)
         container.stop()
+        logging.info(f'Container {container.id} stopped successfully')
 
     # Reboot the node instance
     def reboot_instances(self, node_name):
+        """Restart a container instance (works with both Docker and Podman)."""
+        logging.info(f'Restarting {self.runtime} container for node: {node_name}')
         container = self.client.containers.get(node_name)
         container.restart()
+        logging.info(f'Container {container.id} restarted successfully')
 
     # Terminate the node instance
     def terminate_instances(self, node_name):
+        """Stop and remove a container instance (works with both Docker and Podman)."""
+        logging.info(f'Terminating {self.runtime} container for node: {node_name}')
         container = self.client.containers.get(node_name)
         container.stop()
         container.remove()
+        logging.info(f'Container {container.id} terminated and removed successfully')
 
 
 class docker_node_scenarios(abstract_node_scenarios):
+    """
+    Node chaos scenarios for containerized Kubernetes nodes.
+
+    Supports both Docker and Podman container runtimes. This class provides
+    methods to inject chaos into Kubernetes nodes running as containers
+    (e.g., Kind clusters, Podman-based clusters).
+    """
     def __init__(self, kubecli: KrknKubernetes, node_action_kube_check: bool, affected_nodes_status: AffectedNodeStatus):
+        logging.info('Initializing docker_node_scenarios (supports Docker and Podman)')
         super().__init__(kubecli, node_action_kube_check, affected_nodes_status)
         self.docker = Docker()
         self.node_action_kube_check = node_action_kube_check
+        logging.info(f'Node scenarios initialized successfully using {self.docker.runtime} runtime')
 
     # Node scenario to start the node
     def node_start_scenario(self, instance_kill_count, node, timeout, poll_interval):

requirements.txt

@@ -9,7 +9,7 @@ azure-mgmt-network==27.0.0
 itsdangerous==2.0.1
 coverage==7.6.12
 datetime==5.4
-docker==7.0.0
+docker>=6.0,<7.0  # docker 7.0+ has breaking changes with Unix sockets
 gitpython==3.1.41
 google-auth==2.37.0
 google-cloud-compute==1.22.0
@@ -28,7 +28,8 @@ pyfiglet==1.0.2
 pytest==8.0.0
 python-ipmi==0.5.4
 python-openstackclient==6.5.0
-requests==2.32.4
+requests<2.32  # requests 2.32+ breaks Unix socket support (http+docker scheme)
+requests-unixsocket>=0.4.0  # Required for Docker Unix socket support
 service_identity==24.1.0
 PyYAML==6.0.1
 setuptools==78.1.1

scenarios/kind/node_scenarios_example.yml

@@ -1,16 +1,18 @@
 node_scenarios:
   - actions:                                                        # node chaos scenarios to be injected
     - node_stop_start_scenario
-    node_name: kind-worker                                          # node on which scenario has to be injected; can set multiple names separated by comma
-    # label_selector: node-role.kubernetes.io/worker                # when node_name is not specified, a node with matching label_selector is selected for node chaos scenario injection
+    # node_name: kind-control-plane                                        # node on which scenario has to be injected; can set multiple names separated by comma
+    label_selector: kubernetes.io/hostname=kind-worker              # when node_name is not specified, a node with matching label_selector is selected for node chaos scenario injection
     instance_count: 1                                               # Number of nodes to perform action/select that match the label selector
     runs: 1                                                         # number of times to inject each scenario under actions (will perform on same node each time)
     timeout: 120                                                    # duration to wait for completion of node scenario injection
     cloud_type: docker                                                # cloud type on which Kubernetes/OpenShift runs
+    duration: 10
   - actions:
     - node_reboot_scenario
-    node_name: kind-worker
-    # label_selector: node-role.kubernetes.io/infra
+    node_name: kind-control-plane
+    # label_selector: kubernetes.io/hostname=kind-worker
     instance_count: 1
     timeout: 120
     cloud_type: docker
+    kube_check: false