Use 'v' prefix for default exporter image tag

* feat: Implement iperf3 exporter core logic and log level configuration

This commit completes the core functionality of the iperf3 exporter and adds flexible log level configuration.

Key changes:
- Added command-line (`--log-level`) and environment variable (`LOG_LEVEL`) options to configure the logging level.
- Implemented the main test orchestration loop (`main_loop`) which:
    - Discovers iperf3 server pods via the Kubernetes API.
    - Periodically runs iperf3 tests (TCP/UDP) between the exporter pod and discovered server pods.
    - Avoids self-testing.
    - Uses configurable test intervals, server ports, and protocols.
    - Requires `SOURCE_NODE_NAME` to be set.
- Refined the `parse_and_publish_metrics` function to:
    - Accurately parse iperf3 results for bandwidth, jitter, packets, and lost packets.
    - Set `IPERF_TEST_SUCCESS` metric (0 for failure, 1 for success).
    - Zero out all relevant metrics for a given path upon test failure to prevent stale data.
    - Handle UDP-specific metrics correctly, zeroing them for TCP tests.
    - Improved robustness in accessing iperf3 result attributes.
- Updated the main execution block to initialize logging, start the Prometheus HTTP server, and invoke the main loop.
- Added comprehensive docstrings and inline comments throughout `exporter/exporter.py` for improved readability and maintainability.

These changes align the exporter's implementation with the details specified in the design document (docs/DESIGN.MD).

* feat: Update Helm chart and CI for exporter enhancements

This commit introduces updates to the Helm chart to support log level
configuration for the iperf3 exporter, and modifies the CI workflow
to improve image tagging for pull requests.

Helm Chart Changes (`charts/iperf3-monitor`):
- Added `exporter.logLevel` to `values.yaml` (default: "INFO") to allow
  you to set the exporter's log level.
- Updated `templates/exporter-deployment.yaml` to use the
  `exporter.logLevel` value to set the `LOG_LEVEL` environment
  variable in the exporter container.

CI Workflow Changes (`.github/workflows/ci.yaml`):
- Modified the Docker image build process to tag images built from
  pull requests with `pr-<PR_NUMBER>`.
- Ensured that these PR-specific images are pushed to the container
  registry.
- Preserved existing tagging mechanisms (e.g., SHA-based tags).

* fix: Add Docker login and permissions to CI workflow

This commit fixes the Docker image push failure in the CI workflow
by adding the necessary Docker login step and ensuring the correct
permissions are set for the GITHUB_TOKEN.

- Added a Docker login step using `docker/login-action@v3` to the
  `Build Docker Image` job in `.github/workflows/ci.yaml`. This
  authenticates to GHCR before attempting to push images.
- Added a `permissions` block to the `Build Docker Image` job, granting
  `packages: write` scope to the `GITHUB_TOKEN`. This is required
  to allow pushing packages to the GitHub Container Registry.

---------

Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>

.github/workflows/ci.yaml

@@ -27,23 +27,49 @@ jobs:
   build:
     name: Build Docker Image
     runs-on: ubuntu-latest
+    permissions:
+      contents: read # Needed to checkout the repository
+      packages: write # Needed to push Docker images to GHCR
     steps:
       - name: Check out code
         uses: actions/checkout@v3
 
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v2
+
       - name: Extract metadata (tags, labels) for Docker
         id: meta
         uses: docker/metadata-action@v4
         with:
           images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}
+          tags: |
+            # Tag with the PR number if it's a pull request event
+            type=match,pattern=pull_request,value=pr-{{number}}
+            # Tag with the git SHA
+            type=sha,prefix=
+            # Tag with 'latest' if on the main branch (though this workflow only runs on PRs to main)
+            type=ref,event=branch,pattern=main,value=latest
+
+      - name: Log in to GitHub Container Registry
+        uses: docker/login-action@v3
+        with:
+          registry: ${{ env.REGISTRY }}
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
 
       - name: Build Docker image
         uses: docker/build-push-action@v4
         with:
           context: ./exporter
-          push: false # Do not push on PRs
+          # Push the image if the event is a pull request.
+          # The workflow currently only triggers on pull_request events.
+          push: ${{ github.event_name == 'pull_request' }}
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
+          platforms: linux/amd64,linux/arm64
 
   test:
     name: Run Tests

.github/workflows/release.yml

@@ -36,6 +36,12 @@ jobs:
       - name: Check out code
         uses: actions/checkout@v3
 
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v2
+
       - name: Log in to GitHub Container Registry
         uses: docker/login-action@v2
         with:
@@ -56,6 +62,7 @@ jobs:
           push: true
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
+          platforms: linux/amd64,linux/arm64
 
   package-and-publish-chart:
     name: Package and Publish Helm Chart
@@ -82,8 +89,8 @@ jobs:
       - name: Set Chart Version from Tag
         run: |
           VERSION=$(echo "${{ github.ref_name }}" | sed 's/^v//')
-          yq e -i '.version = strenv(VERSION)' ./charts/iperf3-monitor/Chart.yaml
+          VERSION=$VERSION yq e -i '.version = strenv(VERSION)' ./charts/iperf3-monitor/Chart.yaml
-          yq e -i '.appVersion = strenv(VERSION)' ./charts/iperf3-monitor/Chart.yaml
+          VERSION=$VERSION yq e -i '.appVersion = strenv(VERSION)' ./charts/iperf3-monitor/Chart.yaml
           cat ./charts/iperf3-monitor/Chart.yaml # Optional: print updated Chart.yaml
 
       - name: Publish Helm chart

.gitignore

@@ -36,4 +36,4 @@ Thumbs.db
 
 # Helm
 !charts/iperf3-monitor/.helmignore
-charts/*.tgz # Ignore packaged chart files
+charts/iperf3-monitor/charts/

README.md

@@ -37,7 +37,7 @@ This separation of concerns ensures scalability, resilience, and aligns with Kub
 1.  Add the Helm chart repository (replace with your actual repo URL once published):
 
     ```/dev/null/helm-install.sh#L1-1
-    helm repo add iperf3-monitor https://your-github-org.github.io/iperf3-monitor/
+    helm repo add iperf3-monitor https://malarinv.github.io/iperf3-monitor/
     ```
 
 2.  Update your Helm repositories:
@@ -78,7 +78,7 @@ exporter:
   # -- Configuration for the exporter container image.
   image:
     # -- The container image repository for the exporter.
-    repository: ghcr.io/my-org/iperf3-prometheus-exporter # Replace with your repo URL
+    repository: ghcr.io/malarinv/iperf3-monitor
     # -- The container image tag for the exporter. If not set, the chart's appVersion is used.
     tag: ""
     # -- The image pull policy for the exporter container.
@@ -430,8 +430,4 @@ The project includes a GitHub Actions workflow (`.github/workflows/release.yml`)
 
 ## License
 
-This project is licensed under the terms defined in the `LICENSE` file.
+This project is licensed under the GNU Affero General Public License v3. See the `LICENSE` file for details.
-
-```iperf3-monitor/LICENSE
-This project is currently unlicensed. Please see the project's documentation or repository for licensing information when it becomes available.
-```

charts/iperf3-monitor/Chart.lock

@@ -0,0 +1,9 @@
+dependencies:
+- name: kube-prometheus-stack
+  repository: https://prometheus-community.github.io/helm-charts
+  version: 75.3.6
+- name: prometheus-operator
+  repository: oci://tccr.io/truecharts
+  version: 11.5.1
+digest: sha256:3000e63445f8ba8df601cb483f4f77d14c5c4662bff2d16ffcf5cf1f7def314b
+generated: "2025-06-20T17:25:44.538372209+05:30"

charts/iperf3-monitor/Chart.yaml

@@ -12,19 +12,23 @@ keywords:
   - kubernetes
   - prometheus
   - grafana
-home: https://github.com/malarinv/iperf3-monitor # Replace with your repo URL
+home: https://github.com/malarinv/iperf3-monitor
 sources:
-  - https://github.com/malarinv/iperf3-monitor # Replace with your repo URL
+  - https://github.com/malarinv/iperf3-monitor
 maintainers:
-  - name: Malar Invention # Replace with your name
+  - name: Malar Invention
-    email: malarkannan.invention@gmail.com # Replace with your email
+    email: malarkannan.invention@gmail.com
 icon: https://raw.githubusercontent.com/malarinv/iperf3-monitor/main/icon.png # Optional icon URL
 annotations:
   artifacthub.io/changes: |
     - Add initial Helm chart structure.
   artifacthub.io/category: networking
 dependencies:
-  - name: prometheus-community/kube-prometheus-stack # Example dependency if you package the whole stack
+  - name: kube-prometheus-stack # Example dependency if you package the whole stack
     version: ">=30.0.0" # Specify a compatible version range
     repository: https://prometheus-community.github.io/helm-charts
-    condition: serviceMonitor.enabled # Only include if ServiceMonitor is enabled (assuming Prometheus Operator)
+    condition: "dependencies.install, serviceMonitor.enabled, !dependencies.useTrueChartsPrometheusOperator"
+  - name: prometheus-operator
+    version: ">=8.11.1"
+    repository: "oci://tccr.io/truecharts"
+    condition: "dependencies.install, serviceMonitor.enabled, dependencies.useTrueChartsPrometheusOperator"

charts/iperf3-monitor/templates/_helpers.tpl

@@ -29,9 +29,9 @@ Create chart's labels
 {{- define "iperf3-monitor.labels" -}}
 helm.sh/chart: {{ include "iperf3-monitor.name" . }}-{{ .Chart.Version | replace "+" "_" }}
 {{ include "iperf3-monitor.selectorLabels" . }}
-{{- if .Chart.AppVersion -}}
+{{ if .Chart.AppVersion }}
 app.kubernetes.io/version: {{ .Chart.AppVersion | quote }}
-{{- end -}}
+{{ end }}
 app.kubernetes.io/managed-by: {{ .Release.Service }}
 {{- end -}}
 
@@ -41,13 +41,13 @@ Selector labels
 {{- define "iperf3-monitor.selectorLabels" -}}
 app.kubernetes.io/name: {{ include "iperf3-monitor.name" . }}
 app.kubernetes.io/instance: {{ .Release.Name }}
-{{- end -}}
+{{ end }}
 
 {{/*
 Create the name of the service account to use
 */}}
 {{- define "iperf3-monitor.serviceAccountName" -}}
-{{- if .Values.serviceAccount.create -}}
+{{- if .Values.rbac.create -}}
     {{- default (include "iperf3-monitor.fullname" .) .Values.serviceAccount.name -}}
 {{- else -}}
     {{- default "default" .Values.serviceAccount.name -}}

charts/iperf3-monitor/templates/exporter-deployment.yaml

@@ -20,7 +20,7 @@ spec:
       serviceAccountName: {{ include "iperf3-monitor.serviceAccountName" . }}
       containers:
       - name: iperf3-exporter
-        image: "{{ .Values.exporter.image.repository }}:{{ .Values.exporter.image.tag | default .Chart.AppVersion }}"
+        image: "{{ .Values.exporter.image.repository }}:{{ .Values.exporter.image.tag | default (printf "v%s" .Chart.AppVersion) }}"
         imagePullPolicy: {{ .Values.exporter.image.pullPolicy }}
         ports:
         - containerPort: {{ .Values.service.targetPort }}
@@ -34,6 +34,8 @@ spec:
           value: "{{ .Values.exporter.testInterval }}"
         - name: IPERF_TEST_PROTOCOL
           value: "{{ .Values.exporter.testProtocol }}"
+        - name: LOG_LEVEL
+          value: "{{ .Values.exporter.logLevel }}"
         - name: IPERF_SERVER_PORT
           value: "5201" # Hardcoded as per server DaemonSet
         - name: IPERF_SERVER_NAMESPACE
@@ -41,9 +43,8 @@ spec:
             fieldRef:
               fieldPath: metadata.namespace
         - name: IPERF_SERVER_LABEL_SELECTOR
-          value: "app.kubernetes.io/name={{ include \"iperf3-monitor.name\" . }},app.kubernetes.io/instance={{ .Release.Name }},app.kubernetes.io/component=server"
+          value: 'app.kubernetes.io/name={{ include "iperf3-monitor.name" . }},app.kubernetes.io/instance={{ .Release.Name }},app.kubernetes.io/component=server'
         {{- with .Values.exporter.resources }}
         resources:
           {{- toYaml . | nindent 10 }}
         {{- end }}
-```

charts/iperf3-monitor/templates/server-daemonset.yaml

@@ -42,4 +42,3 @@ spec:
           resources:
             {{- toYaml . | nindent 12 }}
           {{- end }}
-```

charts/iperf3-monitor/values.yaml

@@ -12,7 +12,7 @@ exporter:
   # -- Configuration for the exporter container image.
   image:
     # -- The container image repository for the exporter.
-    repository: ghcr.io/malarinv/iperf3-prometheus-exporter # Replace with your repo URL
+    repository: ghcr.io/malarinv/iperf3-monitor
     # -- The container image tag for the exporter. If not set, the chart's appVersion is used.
     tag: ""
     # -- The image pull policy for the exporter container.
@@ -24,6 +24,9 @@ exporter:
   # -- Interval in seconds between complete test cycles (i.e., testing all server nodes).
   testInterval: 300
 
+  # -- Log level for the iperf3 exporter (e.g., DEBUG, INFO, WARNING, ERROR, CRITICAL).
+  logLevel: INFO
+
   # -- Timeout in seconds for a single iperf3 test run.
   testTimeout: 10
 
@@ -85,7 +88,7 @@ rbac:
 serviceAccount:
   # -- The name of the ServiceAccount to use for the exporter pod.
   # Only used if rbac.create is false. If not set, it defaults to the chart's fullname.
-  name: ""
+  name: "iperf3-monitor"
 
 serviceMonitor:
   # -- If true, create a ServiceMonitor resource for integration with Prometheus Operator.
@@ -118,3 +121,19 @@ networkPolicy:
   namespaceSelector: {}
   # -- Specify pod selectors if needed.
   podSelector: {}
+
+# -----------------------------------------------------------------------------
+# Dependency Configuration
+# -----------------------------------------------------------------------------
+dependencies:
+  # -- Set to true to install Prometheus operator dependency if serviceMonitor.enabled is also true.
+  # -- Set to false to disable the installation of Prometheus operator dependency,
+  # -- regardless of serviceMonitor.enabled. This is useful if you have Prometheus
+  # -- Operator installed and managed separately in your cluster.
+  install: true
+
+  # -- Set to true to use the TrueCharts Prometheus Operator instead of kube-prometheus-stack.
+  # This chart's ServiceMonitor resources require a Prometheus Operator to be functional.
+  # If serviceMonitor.enabled is true and dependencies.install is true,
+  # one of these two dependencies will be pulled based on this flag.
+  useTrueChartsPrometheusOperator: false

devbox.json

@@ -0,0 +1,14 @@
+{
+  "$schema":  "https://raw.githubusercontent.com/jetify-com/devbox/0.13.7/.schema/devbox.schema.json",
+  "packages": [],
+  "shell": {
+    "init_hook": [
+      "echo 'Welcome to devbox!' > /dev/null"
+    ],
+    "scripts": {
+      "test": [
+        "echo \"Error: no test specified\" && exit 1"
+      ]
+    }
+  }
+}

devbox.lock

@@ -0,0 +1,4 @@
+{
+  "lockfile_version": "1",
+  "packages": {}
+}

exporter/Dockerfile

@@ -1,11 +1,15 @@
 # Stage 1: Build stage with dependencies
 FROM python:3.9-slim as builder
 
+# Declare TARGETARCH for use in this stage
+ARG TARGETARCH
 WORKDIR /app
 
-# Install iperf3 and build dependencies
+# Minimal dependencies for builder stage if any Python packages had C extensions.
+# Assuming requirements.txt does not need gcc or other build tools for now.
+# If pip install fails later, add necessary build tools (e.g., gcc, python3-dev) here.
 RUN apt-get update && \
-    apt-get install -y --no-install-recommends gcc iperf3 libiperf-dev && \
+    # apt-get install -y --no-install-recommends gcc python3-dev # Example if needed
     rm -rf /var/lib/apt/lists/*
 
 # Install Python dependencies
@@ -17,9 +21,11 @@ FROM python:3.9-slim
 
 WORKDIR /app
 
-# Copy iperf3 binary and library from the builder stage
+# Install iperf3 and its runtime dependency libsctp1 directly in the final stage.
-COPY --from=builder /usr/bin/iperf3 /usr/bin/iperf3
+# This simplifies the Dockerfile by removing the need to copy iperf3 components from the builder.
-COPY --from=builder /usr/lib/x86_64-linux-gnu/libiperf.so.0 /usr/lib/x86_64-linux-gnu/libiperf.so.0
+RUN apt-get update && \
+    apt-get install -y --no-install-recommends iperf3 libsctp1 && \
+    rm -rf /var/lib/apt/lists/*
 
 # Copy installed Python packages from the builder stage
 COPY --from=builder /usr/local/lib/python3.9/site-packages /usr/local/lib/python3.9/site-packages

exporter/exporter.py

@@ -1,28 +1,60 @@
+"""
+Prometheus exporter for iperf3 network performance monitoring.
+
+This script runs iperf3 tests between the node it's running on (source) and
+other iperf3 server pods discovered in a Kubernetes cluster. It then exposes
+these metrics for Prometheus consumption.
+
+Configuration is primarily through environment variables and command-line arguments
+for log level.
+"""
 import os
 import time
 import logging
+import argparse
+import sys
 from kubernetes import client, config
 from prometheus_client import start_http_server, Gauge
 import iperf3
 
-# --- Configuration ---
+# --- Global Configuration & Setup ---
-# Configure logging
+
-logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
+# Argument parsing for log level configuration
+# The command-line --log-level argument takes precedence over the LOG_LEVEL env var.
+# Defaults to INFO if neither is set.
+parser = argparse.ArgumentParser(description="iperf3 Prometheus exporter.")
+parser.add_argument(
+    '--log-level',
+    default=os.environ.get('LOG_LEVEL', 'INFO').upper(),
+    choices=['DEBUG', 'INFO', 'WARNING', 'ERROR', 'CRITICAL'],
+    help='Set the logging level. Overrides LOG_LEVEL environment variable. (Default: INFO)'
+)
+args = parser.parse_args()
+log_level_str = args.log_level
+
+# Convert log level string (e.g., 'INFO') to its numeric representation (e.g., logging.INFO)
+numeric_level = getattr(logging, log_level_str.upper(), None)
+if not isinstance(numeric_level, int):
+    # This case should ideally not be reached if choices in argparse are respected.
+    logging.error(f"Invalid log level: {log_level_str}. Defaulting to INFO.")
+    numeric_level = logging.INFO
+logging.basicConfig(level=numeric_level, format='%(asctime)s - %(levelname)s - %(message)s')
 
 # --- Prometheus Metrics Definition ---
+# These gauges will be used to expose iperf3 test results.
 IPERF_BANDWIDTH_MBPS = Gauge(
     'iperf_network_bandwidth_mbps',
-    'Network bandwidth measured by iperf3 in Megabits per second',
+    'Network bandwidth measured by iperf3 in Megabits per second (Mbps)',
     ['source_node', 'destination_node', 'protocol']
 )
 IPERF_JITTER_MS = Gauge(
     'iperf_network_jitter_ms',
-    'Network jitter measured by iperf3 in milliseconds',
+    'Network jitter measured by iperf3 in milliseconds (ms) for UDP tests',
     ['source_node', 'destination_node', 'protocol']
 )
 IPERF_PACKETS_TOTAL = Gauge(
     'iperf_network_packets_total',
-    'Total packets transmitted or received during the iperf3 test',
+    'Total packets transmitted/received during the iperf3 UDP test',
     ['source_node', 'destination_node', 'protocol']
 )
 IPERF_LOST_PACKETS = Gauge(
@@ -38,12 +70,21 @@ IPERF_TEST_SUCCESS = Gauge(
 
 def discover_iperf_servers():
     """
-    Discover iperf3 server pods in the cluster using the Kubernetes API.
+    Discovers iperf3 server pods within a Kubernetes cluster.
+
+    It uses the in-cluster Kubernetes configuration to connect to the API.
+    The target namespace and label selector for iperf3 server pods are configured
+    via environment variables:
+    - IPERF_SERVER_NAMESPACE (default: 'default')
+    - IPERF_SERVER_LABEL_SELECTOR (default: 'app=iperf3-server')
+
+    Returns:
+        list: A list of dictionaries, where each dictionary contains the 'ip'
+              and 'node_name' of a discovered iperf3 server pod. Returns an
+              empty list if discovery fails or no servers are found.
     """
     try:
-        # Load in-cluster configuration
+        config.load_incluster_config() # Assumes running inside a Kubernetes pod
-        # Assumes the exporter runs in a pod with a service account having permissions
-        config.load_incluster_config()
         v1 = client.CoreV1Api()
 
         namespace = os.getenv('IPERF_SERVER_NAMESPACE', 'default')
@@ -51,110 +92,206 @@ def discover_iperf_servers():
 
         logging.info(f"Discovering iperf3 servers with label '{label_selector}' in namespace '{namespace}'")
 
-        # List pods across all namespaces with the specified label selector
-        # Note: list_pod_for_all_namespaces requires cluster-wide permissions
         ret = v1.list_pod_for_all_namespaces(label_selector=label_selector, watch=False)
 
         servers = []
-        for i in ret.items:
+        for item in ret.items:
-            # Ensure pod has an IP and is running
+            if item.status.pod_ip and item.status.phase == 'Running':
-            if i.status.pod_ip and i.status.phase == 'Running':
                 servers.append({
-                    'ip': i.status.pod_ip,
+                    'ip': item.status.pod_ip,
-                    'node_name': i.spec.node_name
+                    'node_name': item.spec.node_name # Node where the iperf server pod is running
                 })
         logging.info(f"Discovered {len(servers)} iperf3 server pods.")
         return servers
+    except config.ConfigException as e:
+        logging.error(f"Kubernetes config error: {e}. Is the exporter running in a cluster with RBAC permissions?")
+        return []
     except Exception as e:
         logging.error(f"Error discovering iperf servers: {e}")
-        return [] # Return empty list on error to avoid crashing the loop
+        return [] # Return empty list on error to avoid crashing the main loop
 
-def run_iperf_test(server_ip, server_port, protocol, source_node, dest_node):
+def run_iperf_test(server_ip, server_port, protocol, source_node_name, dest_node_name):
     """
-    Runs a single iperf3 test and updates Prometheus metrics.
+    Runs a single iperf3 test against a specified server and publishes metrics.
+
+    Args:
+        server_ip (str): The IP address of the iperf3 server.
+        server_port (int): The port number of the iperf3 server.
+        protocol (str): The protocol to use ('tcp' or 'udp').
+        source_node_name (str): The name of the source node (where this exporter is running).
+        dest_node_name (str): The name of the destination node (where the server is running).
+
+    The test duration is controlled by the IPERF_TEST_DURATION environment variable
+    (default: 5 seconds).
     """
-    logging.info(f"Running iperf3 test from {source_node} to {dest_node} ({server_ip}:{server_port}) using {protocol.upper()}")
+    logging.info(f"Running iperf3 {protocol.upper()} test from {source_node_name} to {dest_node_name} ({server_ip}:{server_port})")
 
-    client = iperf3.Client()
+    iperf_client = iperf3.Client()
-    client.server_hostname = server_ip
+    iperf_client.server_hostname = server_ip
-    client.port = server_port
+    iperf_client.port = server_port
-    client.protocol = protocol
+    iperf_client.protocol = protocol
-    # Duration of the test (seconds)
+    iperf_client.duration = int(os.getenv('IPERF_TEST_DURATION', 5)) # Test duration in seconds
-    client.duration = int(os.getenv('IPERF_TEST_DURATION', 5))
+    iperf_client.json_output = True # Enables easy parsing of results
-    # Output results as JSON for easy parsing
-    client.json_output = True
 
-    result = client.run()
+    try:
+        result = iperf_client.run()
+        parse_and_publish_metrics(result, source_node_name, dest_node_name, protocol)
+    except Exception as e:
+        # Catch unexpected errors during client.run() or parsing
+        logging.error(f"Exception during iperf3 test or metric parsing for {dest_node_name}: {e}")
+        labels = {'source_node': source_node_name, 'destination_node': dest_node_name, 'protocol': protocol}
+        IPERF_TEST_SUCCESS.labels(**labels).set(0)
+        try:
+            IPERF_BANDWIDTH_MBPS.labels(**labels).set(0)
+            IPERF_JITTER_MS.labels(**labels).set(0)
+            IPERF_PACKETS_TOTAL.labels(**labels).set(0)
+            IPERF_LOST_PACKETS.labels(**labels).set(0)
+        except KeyError:
+            logging.debug(f"KeyError setting failure metrics for {labels} after client.run() exception.")
 
-    # Parse results and update metrics
-    parse_and_publish_metrics(result, source_node, dest_node, protocol)
 
 def parse_and_publish_metrics(result, source_node, dest_node, protocol):
     """
-    Parses the iperf3 result and updates Prometheus gauges.
+    Parses the iperf3 test result and updates Prometheus gauges.
-    Handles both successful and failed tests.
+
+    Args:
+        result (iperf3.TestResult): The result object from the iperf3 client.
+        source_node (str): Name of the source node.
+        dest_node (str): Name of the destination node.
+        protocol (str): Protocol used for the test ('tcp' or 'udp').
     """
     labels = {'source_node': source_node, 'destination_node': dest_node, 'protocol': protocol}
 
-    if result and result.error:
+    # Handle failed tests (e.g., server unreachable) or missing result object
-        logging.error(f"Test from {source_node} to {dest_node} failed: {result.error}")
+    if not result or result.error:
+        error_message = result.error if result and result.error else "No result object from iperf3 client"
+        logging.warning(f"Test from {source_node} to {dest_node} ({protocol.upper()}) failed: {error_message}")
         IPERF_TEST_SUCCESS.labels(**labels).set(0)
-        # Set metrics to 0 on failure
+        # Set all relevant metrics to 0 on failure to clear stale values from previous successes
         try:
             IPERF_BANDWIDTH_MBPS.labels(**labels).set(0)
-            IPERF_JITTER_MS.labels(**labels).set(0)
+            IPERF_JITTER_MS.labels(**labels).set(0) # Applicable for UDP, zeroed for TCP later
-            IPERF_PACKETS_TOTAL.labels(**labels).set(0)
+            IPERF_PACKETS_TOTAL.labels(**labels).set(0) # Applicable for UDP, zeroed for TCP later
-            IPERF_LOST_PACKETS.labels(**labels).set(0)
+            IPERF_LOST_PACKETS.labels(**labels).set(0) # Applicable for UDP, zeroed for TCP later
         except KeyError:
-             # Labels might not be registered yet if this is the first failure
+            # This can happen if labels were never registered due to continuous failures
-             pass
+            logging.debug(f"KeyError when setting failure metrics for {labels}. Gauges might not be initialized.")
         return
 
-    if not result:
+    # If we reach here, the test itself was successful in execution
-         logging.error(f"Test from {source_node} to {dest_node} failed to return a result object.")
-         IPERF_TEST_SUCCESS.labels(**labels).set(0)
-         try:
-            IPERF_BANDWIDTH_MBPS.labels(**labels).set(0)
-            IPERF_JITTER_MS.labels(**labels).set(0)
-            IPERF_PACKETS_TOTAL.labels(**labels).set(0)
-            IPERF_LOST_PACKETS.labels(**labels).set(0)
-         except KeyError:
-             pass
-         return
-
-
     IPERF_TEST_SUCCESS.labels(**labels).set(1)
 
-    # The summary data is typically in result.json['end']['sum_sent'] or result.json['end']['sum_received']
+    # Determine bandwidth:
-    # The iperf3-python client often exposes this directly as attributes like sent_Mbps or received_Mbps
+    # Order of preference: received_Mbps, sent_Mbps, Mbps, then JSON fallbacks.
-    # For TCP, we usually care about the received bandwidth on the client side (which is the exporter)
+    # received_Mbps is often most relevant for TCP client perspective.
-    # For UDP, the client report contains jitter, lost packets, etc.
+    # sent_Mbps can be relevant for UDP or as a TCP fallback.
     bandwidth_mbps = 0
     if hasattr(result, 'received_Mbps') and result.received_Mbps is not None:
         bandwidth_mbps = result.received_Mbps
     elif hasattr(result, 'sent_Mbps') and result.sent_Mbps is not None:
-        # Fallback, though received_Mbps is usually more relevant for TCP client
         bandwidth_mbps = result.sent_Mbps
-    # Add a check for the raw JSON output structure as a fallback
+    elif hasattr(result, 'Mbps') and result.Mbps is not None: # General attribute from iperf3 library
-    elif result.json and 'end' in result.json and 'sum_received' in result.json['end'] and result.json['end']['sum_received']['bits_per_second'] is not None:
+        bandwidth_mbps = result.Mbps
-         bandwidth_mbps = result.json['end']['sum_received']['bits_per_second'] / 1000000
+    # Fallback to raw JSON if direct attributes are None or missing
-    elif result.json and 'end' in result.json and 'sum_sent' in result.json['end'] and result.json['end']['sum_sent']['bits_per_second'] is not None:
+    elif result.json:
-         bandwidth_mbps = result.json['end']['sum_sent']['bits_per_second'] / 1000000
+        # Prefer received sum, then sent sum from the JSON output's 'end' summary
-
+        if 'end' in result.json and 'sum_received' in result.json['end'] and \
+           result.json['end']['sum_received'].get('bits_per_second') is not None:
+            bandwidth_mbps = result.json['end']['sum_received']['bits_per_second'] / 1000000.0
+        elif 'end' in result.json and 'sum_sent' in result.json['end'] and \
+             result.json['end']['sum_sent'].get('bits_per_second') is not None:
+            bandwidth_mbps = result.json['end']['sum_sent']['bits_per_second'] / 1000000.0
 
     IPERF_BANDWIDTH_MBPS.labels(**labels).set(bandwidth_mbps)
 
     # UDP specific metrics
     if protocol == 'udp':
-        # iperf3-python exposes UDP results directly
+        # These attributes are specific to UDP tests in iperf3
-        IPERF_JITTER_MS.labels(**labels).set(result.jitter_ms if hasattr(result, 'jitter_ms') and result.jitter_ms is not None else 0)
+        IPERF_JITTER_MS.labels(**labels).set(getattr(result, 'jitter_ms', 0) if result.jitter_ms is not None else 0)
-        IPERF_PACKETS_TOTAL.labels(**labels).set(result.packets if hasattr(result, 'packets') and result.packets is not None else 0)
+        IPERF_PACKETS_TOTAL.labels(**labels).set(getattr(result, 'packets', 0) if result.packets is not None else 0)
-        IPERF_LOST_PACKETS.labels(**labels).set(result.lost_packets if hasattr(result, 'lost_packets') and result.lost_packets is not None else 0)
+        IPERF_LOST_PACKETS.labels(**labels).set(getattr(result, 'lost_packets', 0) if result.lost_packets is not None else 0)
     else:
-        # Ensure UDP metrics are zeroed or absent for TCP tests
+        # For TCP tests, ensure UDP-specific metrics are set to 0
         try:
              IPERF_JITTER_MS.labels(**labels).set(0)
              IPERF_PACKETS_TOTAL.labels(**labels).set(0)
              IPERF_LOST_PACKETS.labels(**labels).set(0)
         except KeyError:
+            # Can occur if labels not yet registered (e.g. first test is TCP)
+            logging.debug(f"KeyError for {labels} when zeroing UDP metrics for TCP test.")
             pass
+
+def main_loop():
+    """
+    Main operational loop of the iperf3 exporter.
+
+    This loop periodically:
+    1. Fetches configuration from environment variables:
+       - IPERF_TEST_INTERVAL (default: 300s): Time between test cycles.
+       - IPERF_SERVER_PORT (default: 5201): Port for iperf3 servers.
+       - IPERF_TEST_PROTOCOL (default: 'tcp'): 'tcp' or 'udp'.
+       - SOURCE_NODE_NAME (critical): Name of the node this exporter runs on.
+    2. Discovers iperf3 server pods in the Kubernetes cluster.
+    3. Runs iperf3 tests against each discovered server (unless it's on the same node).
+    4. Sleeps for the configured test interval.
+
+    If SOURCE_NODE_NAME is not set, the script will log an error and exit.
+    """
+    # Fetch operational configuration from environment variables
+    test_interval = int(os.getenv('IPERF_TEST_INTERVAL', 300))
+    server_port = int(os.getenv('IPERF_SERVER_PORT', 5201))
+    protocol = os.getenv('IPERF_TEST_PROTOCOL', 'tcp').lower() # Ensure lowercase
+    source_node_name = os.getenv('SOURCE_NODE_NAME')
+
+    # SOURCE_NODE_NAME is crucial for labeling metrics correctly.
+    if not source_node_name:
+        logging.error("CRITICAL: SOURCE_NODE_NAME environment variable not set. This is required. Exiting.")
+        sys.exit(1)
+
+    logging.info(
+        f"Exporter configured. Source Node: {source_node_name}, "
+        f"Test Interval: {test_interval}s, Server Port: {server_port}, Protocol: {protocol.upper()}"
+    )
+
+    while True:
+        logging.info("Starting new iperf test cycle...")
+        servers = discover_iperf_servers()
+
+        if not servers:
+            logging.warning("No iperf servers discovered in this cycle. Check K8s setup and RBAC permissions.")
+        else:
+            for server in servers:
+                dest_node_name = server.get('node_name', 'unknown_destination_node') # Default if key missing
+                server_ip = server.get('ip')
+
+                if not server_ip:
+                    logging.warning(f"Discovered server entry missing an IP: {server}. Skipping.")
+                    continue
+
+                # Avoid testing a node against itself
+                if dest_node_name == source_node_name:
+                    logging.info(f"Skipping test to self: {source_node_name} to {server_ip} (on same node: {dest_node_name}).")
+                    continue
+
+                run_iperf_test(server_ip, server_port, protocol, source_node_name, dest_node_name)
+
+        logging.info(f"Test cycle completed. Sleeping for {test_interval} seconds.")
+        time.sleep(test_interval)
+
+if __name__ == '__main__':
+    # Initial logging (like log level) is configured globally at the start of the script.
+
+    # Fetch Prometheus exporter listen port from environment variable
+    listen_port = int(os.getenv('LISTEN_PORT', 9876))
+
+    try:
+        # Start the Prometheus HTTP server to expose metrics.
+        start_http_server(listen_port)
+        logging.info(f"Prometheus exporter listening on port {listen_port}")
+    except Exception as e:
+        logging.error(f"Failed to start Prometheus HTTP server on port {listen_port}: {e}")
+        sys.exit(1) # Exit if the metrics server cannot start
+
+    # Enter the main operational loop.
+    # main_loop() contains its own critical checks (e.g., SOURCE_NODE_NAME) and will exit if necessary.
+    main_loop()