Merge branch 'main' into fix-integration

.github/workflows/check.yaml

@@ -26,7 +26,7 @@ jobs:
       fail-fast: false
       matrix:
         python-version: ["3.8", "3.10"]
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-22.04
     steps:
     - uses: actions/checkout@v4
       with:
@@ -53,7 +53,7 @@ jobs:
       fail-fast: false
       matrix:
         python-version: ["3.8", "3.10"]
-    runs-on: ubuntu-latest
+    runs-on: ubuntu-22.04
     steps:
     - uses: actions/checkout@v4
       with:
@@ -83,19 +83,19 @@ jobs:
       fail-fast: false
       matrix:
         runs-on: [[ubuntu-22.04], [Ubuntu_ARM64_4C_16G_01]]
-        test-command: ['tox -e func -- --series focal --keep-models', 'tox -e func -- --series jammy --keep-models']
+        test-command: ['tox -e func -- -v --series focal --keep-models', 'tox -e func -- -v --series jammy --keep-models']
         juju-channel: ["3.4/stable"]
     steps:
 
       - uses: actions/checkout@v4
         with:
           submodules: true
 
-        # arm64 runners don't have make or gcc installed by default
+        # arm64 runners don't have gcc installed by default
       - name: Install dependencies
         run: |
           sudo apt update
-          sudo apt install -y make gcc
+          sudo apt install -y gcc
 
       - name: Setup Python
         uses: actions/setup-python@v5
@@ -116,8 +116,22 @@ jobs:
             echo "TEST_MODEL_CONSTRAINTS=arch=arm64" >> "$GITHUB_ENV"
           fi
 
+      - name: Build the charm
+        run: charmcraft -v pack
+
       - name: Run tests
-        run: ${{ matrix.test-command }}
+        run: |
+          # These variables are for a consistent method to find the charm file(s) across all projects.
+          # It is designed to work both with charms that output one file per base,
+          # and charms that output a single file to run on all bases.
+          # Not all charms will use them, and for some charms the variables will resolve to the same file.
+          export CHARM_PATH_NOBLE="$(pwd)/$(ls | grep '.*24.04.*\.charm$')"
+          echo "$CHARM_PATH_NOBLE"
+          export CHARM_PATH_JAMMY="$(pwd)/$(ls | grep '.*22.04.*\.charm$')"
+          echo "$CHARM_PATH_JAMMY"
+          export CHARM_PATH_FOCAL="$(pwd)/$(ls | grep '.*20.04.*\.charm$')"
+          echo "$CHARM_PATH_FOCAL"
+          ${{ matrix.test-command }}
         env:
           TEST_JUJU3: "1"  # https://github.com/openstack-charmers/zaza/pull/653
           TEST_JUJU_CHANNEL: ${{ matrix.juju-channel }}

.github/workflows/release.yaml

@@ -22,7 +22,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        runs-on: [[ubuntu-22.04]]
+        runs-on: [[ubuntu-22.04], [Ubuntu_ARM64_4C_16G_01]]
     steps:
       - name: Checkout
         uses: actions/checkout@v4

.gitignore

@@ -1,48 +1,45 @@
-# This is a template `.gitignore` file for ops charms
-# This file is managed by bootstack-charms-spec and should not be modified
-# within individual charm repos. https://launchpad.net/bootstack-charms-spec
+# This file is centrally managed as a template file in https://github.com/canonical/solutions-engineering-automation
+# To update the file:
+# - Edit it in the canonical/solutions-engineering-automation repository.
+# - Open a PR with the changes.
+# - When the PR merges, the soleng-terraform bot will open a PR to the target repositories with the changes.
 
-# Juju files
-.unit-state.db
-
-# Byte-compiled / optimized / DLL files
+# Python Byte-compiled / optimized / DLL files
 __pycache__/
 *.py[cod]
 *$py.class
 
-# Tests files and dir
+# Test files and directories
 .pytest_cache/
 .coverage
 .tox
-.venv
 reports/
 **/report/
 htmlcov/
 .mypy_cache
 
-# Log files
-*.log
-
-# IDEs
-.idea/
-.vscode/
-
-# vi
-.*.swp
-
-# version data
-repo-info
-version
+# python virtual environments (for local dev)
+.venv
+venv
+env
 
-# Python builds
+# Build artefacts
+output/
+.build/
+build/
+*.charm
+*.snap
+# python build artefacts
 deb_dist/
 dist/
 *.egg-info/
 
-# Snaps
-*.snap
+# Log files
+*.log
 
-# Builds
-.build/
-build/
-*.charm
+# general backup files
+*~
+*.bak
+
+# Note: for editor-specific files, please don't add them here, as they are specific to your environment, not the project.
+# Instead, consider using a global gitignore on your workstation.

config.yaml

@@ -15,6 +15,12 @@ options:
     description: |
       Start the prometheus smartctl exporter at "smartctl-exporter-port". By default,
       it will start at port 10201.
+  dcgm-snap-channel:
+    type: string
+    default: "latest/stable"
+    description: |
+      Channel to install the DCGM snap if the hardware has NVIDIA GPU. By default, it will install
+      from latest/stable
   exporter-log-level:
     type: string
     default: "INFO"

src/charm.py

@@ -13,7 +13,7 @@
 from ops.model import ActiveStatus, BlockedStatus, MaintenanceStatus
 
 from hw_tools import HWTool, HWToolHelper, detect_available_tools
-from service import BaseExporter, ExporterError, HardwareExporter, SmartCtlExporter
+from service import BaseExporter, DCGMExporter, ExporterError, HardwareExporter, SmartCtlExporter
 
 logger = logging.getLogger(__name__)
 
@@ -37,6 +37,7 @@ def __init__(self, *args: Any) -> None:
             metrics_endpoints=[
                 {"path": "/metrics", "port": int(self.model.config["hardware-exporter-port"])},
                 {"path": "/metrics", "port": int(self.model.config["smartctl-exporter-port"])},
+                {"path": "/metrics", "port": 9400},
             ],
             # Setting scrape_timeout as collect_timeout in the `duration` format specified in
             # https://prometheus.io/docs/prometheus/latest/configuration/configuration/#duration
@@ -82,6 +83,9 @@ def exporters(self) -> List[BaseExporter]:
         if stored_tools & SmartCtlExporter.hw_tools():
             exporters.append(SmartCtlExporter(self.charm_dir, self.model.config))
 
+        if stored_tools & DCGMExporter.hw_tools():
+            exporters.append(DCGMExporter(self.model.config))
+
         return exporters
 
     def get_stored_tools(self) -> Set[HWTool]:
@@ -226,7 +230,7 @@ def _on_config_changed(self, event: EventBase) -> None:
                 self.model.unit.status = BlockedStatus(message)
                 return
             for exporter in self.exporters:
-                success = exporter.render_config()
+                success = exporter.configure()
                 if success:
                     exporter.restart()
                 else:

src/hw_tools.py

@@ -18,6 +18,7 @@
 import requests
 import urllib3
 from charms.operator_libs_linux.v0 import apt
+from charms.operator_libs_linux.v2 import snap
 from ops.model import ModelError, Resources
 
 import apt_helpers
@@ -187,6 +188,57 @@ def remove(self) -> None:
         """Remove details."""
 
 
+class SnapStrategy(StrategyABC):
+    """Snap strategy class."""
+
+    channel: str
+
+    @property
+    def snap(self) -> str:
+        """Snap name."""
+        return self._name.value
+
+    def install(self) -> None:
+        """Install the snap from a channel."""
+        try:
+            snap.add(self.snap, channel=self.channel)
+            logger.info("Installed %s from channel: %s", self.snap, self.channel)
+
+        # using the snap.SnapError will result into:
+        # TypeError: catching classes that do not inherit from BaseException is not allowed
+        except Exception as err:  # pylint: disable=broad-except
+            logger.error("Failed to install %s from channel: %s: %s", self.snap, self.channel, err)
+            raise err
+
+    def remove(self) -> None:
+        """Remove the snap."""
+        try:
+            snap.remove([self.snap])
+
+        # using the snap.SnapError will result into:
+        # TypeError: catching classes that do not inherit from BaseException is not allowed
+        except Exception as err:  # pylint: disable=broad-except
+            logger.error("Failed to remove %s: %s", self.snap, err)
+            raise err
+
+    def check(self) -> bool:
+        """Check if all services are active."""
+        return all(
+            service.get("active", False)
+            for service in snap.SnapCache()[self.snap].services.values()
+        )
+
+
+class DCGMExporterStrategy(SnapStrategy):
+    """DCGM strategy class."""
+
+    _name = HWTool.DCGM
+
+    def __init__(self, channel: str) -> None:
+        """Init."""
+        self.channel = channel
+
+
 class StorCLIStrategy(TPRStrategyABC):
     """Strategy to install storcli."""
 
@@ -689,13 +741,12 @@ def install(self, resources: Resources, hw_available: Set[HWTool]) -> Tuple[bool
             if strategy.name not in hw_available:
                 continue
             try:
-                # TPRStrategy
                 if isinstance(strategy, TPRStrategyABC):
                     path = fetch_tools.get(strategy.name)  # pylint: disable=W0212
                     if path:
                         strategy.install(path)
-                # APTStrategy
-                elif isinstance(strategy, APTStrategyABC):
+
+                elif isinstance(strategy, (APTStrategyABC, SnapStrategy)):
                     strategy.install()  # pylint: disable=E1120
                 logger.info("Strategy %s install success", strategy)
             except (
@@ -717,7 +768,7 @@ def remove(self, resources: Resources, hw_available: Set[HWTool]) -> None:
         for strategy in self.strategies:
             if strategy.name not in hw_available:
                 continue
-            if isinstance(strategy, (TPRStrategyABC, APTStrategyABC)):
+            if isinstance(strategy, (TPRStrategyABC, APTStrategyABC, SnapStrategy)):
                 strategy.remove()
             logger.info("Strategy %s remove success", strategy)