0.8 - fix health check, added pod counts on node

.gitignore

@@ -190,3 +190,4 @@ pyrightconfig.json
 
 # End of https://www.toptal.com/developers/gitignore/api/python,linux
 myvalues.yaml
+

app.py

@@ -1,5 +1,6 @@
 import os
 import configparser
+from concurrent.futures import ThreadPoolExecutor
 from flask import Flask, jsonify, render_template
 from kubernetes import client, config
 from kubernetes.client.rest import ApiException
@@ -17,7 +18,13 @@ def get_version():
 
 app = Flask(__name__)
 
+_v1_client = None
+_custom_client = None
+
 def get_k8s_client():
+    global _v1_client, _custom_client
+    if _v1_client is not None and _custom_client is not None:
+        return _v1_client, _custom_client
     try:
         config.load_incluster_config()
     except config.ConfigException:
@@ -36,7 +43,9 @@ def get_k8s_client():
                     pass
             if not loaded:
                 raise RuntimeError(f"Failed to load kubeconfig. Make sure the file exists and is accessible. Error: {e}")
-    return client.CoreV1Api(), client.CustomObjectsApi()
+    _v1_client = client.CoreV1Api()
+    _custom_client = client.CustomObjectsApi()
+    return _v1_client, _custom_client
 
 def parse_cpu(cpu_str):
     if not cpu_str: return 0
@@ -64,32 +73,62 @@ def parse_memory(mem_str):
     except:
         return 0
 
+K8S_TIMEOUT = 5.0
+METRICS_TIMEOUT = 3.0
+
+def fetch_nodes(v1):
+    return v1.list_node(_request_timeout=K8S_TIMEOUT).items
+
+def fetch_pods(v1):
+    return v1.list_pod_for_all_namespaces(_request_timeout=K8S_TIMEOUT).items
+
+def fetch_node_metrics(custom):
+    node_metrics = {}
+    try:
+        nm = custom.list_cluster_custom_object("metrics.k8s.io", "v1beta1", "nodes", _request_timeout=METRICS_TIMEOUT)
+        for item in nm.get('items', []):
+            node_metrics[item['metadata']['name']] = item['usage']
+    except Exception as e:
+        print(f"Metrics server node fetch error or unavailable: {e}")
+    return node_metrics
+
+def fetch_pod_metrics(custom):
+    pod_metrics = {}
+    try:
+        pm = custom.list_cluster_custom_object("metrics.k8s.io", "v1beta1", "pods", _request_timeout=METRICS_TIMEOUT)
+        for item in pm.get('items', []):
+            key = f"{item['metadata']['namespace']}/{item['metadata']['name']}"
+            pod_metrics[key] = item['containers']
+    except Exception as e:
+        print(f"Metrics server pod fetch error or unavailable: {e}")
+    return pod_metrics
+
 @app.route('/api/cluster-data')
 def cluster_data():
-    v1, custom = get_k8s_client()
-    
-    cluster = {
-        "name": "Cluster",
-        "children": []
-    }
-    
     try:
-        nodes = v1.list_node().items
+        v1, custom = get_k8s_client()
-        pods = v1.list_pod_for_all_namespaces().items
+
-        
+        with ThreadPoolExecutor(max_workers=4) as executor:
-        node_metrics = {}
+            fut_nodes = executor.submit(fetch_nodes, v1)
-        pod_metrics = {}
+            fut_pods = executor.submit(fetch_pods, v1)
-        try:
+            fut_node_metrics = executor.submit(fetch_node_metrics, custom)
-            nm = custom.list_cluster_custom_object("metrics.k8s.io", "v1beta1", "nodes")
+            fut_pod_metrics = executor.submit(fetch_pod_metrics, custom)
-            for item in nm.get('items', []):
+
-                node_metrics[item['metadata']['name']] = item['usage']
+            nodes = fut_nodes.result()
-                
+            pods = fut_pods.result()
-            pm = custom.list_cluster_custom_object("metrics.k8s.io", "v1beta1", "pods")
+            node_metrics = fut_node_metrics.result()
-            for item in pm.get('items', []):
+            pod_metrics = fut_pod_metrics.result()
-                key = f"{item['metadata']['namespace']}/{item['metadata']['name']}"
+
-                pod_metrics[key] = item['containers']
+        cluster = {
-        except ApiException:
+            "name": "Cluster",
-            pass # Metrics server not available
+            "children": []
+        }
+
+        pods_by_node = {}
+        for pod in pods:
+            nname = pod.spec.node_name
+            if nname:
+                pods_by_node.setdefault(nname, []).append(pod)
             
         for node in nodes:
             node_name = node.metadata.name
@@ -131,6 +170,13 @@ def cluster_data():
             node_usage_cpu = parse_cpu(node_metrics.get(node_name, {}).get('cpu', '0'))
             node_usage_mem = parse_memory(node_metrics.get(node_name, {}).get('memory', '0'))
 
+            node_pods = pods_by_node.get(node_name, [])
+            non_terminated_pods = [
+                p for p in node_pods
+                if p.status and getattr(p.status, 'phase', None) not in ['Succeeded', 'Failed']
+            ]
+            non_terminated_pod_count = len(non_terminated_pods)
+
             node_data = {
                 "name": node_name,
                 "type": "node",
@@ -141,6 +187,10 @@ def cluster_data():
                 "internalIp": internal_ip,
                 "os_image": os_image,
                 "osImage": os_image,
+                "non_terminated_pods": non_terminated_pod_count,
+                "nonTerminatedPods": non_terminated_pod_count,
+                "pod_count": non_terminated_pod_count,
+                "podCount": non_terminated_pod_count,
                 "cpu_allocatable": cpu_allocatable,
                 "mem_allocatable": mem_allocatable,
                 "cpu_usage": node_usage_cpu,
@@ -151,42 +201,41 @@ def cluster_data():
             node_cpu_total = 0
             node_mem_total = 0
             
-            for pod in pods:
+            for pod in node_pods:
-                if pod.spec.node_name == node_name:
+                pod_cpu_req = 0
-                    pod_cpu_req = 0
+                pod_mem_req = 0
-                    pod_mem_req = 0
+                
+                if pod.spec.containers:
+                    for c in pod.spec.containers:
+                        if c.resources and c.resources.requests:
+                            pod_cpu_req += parse_cpu(c.resources.requests.get('cpu', '0'))
+                            pod_mem_req += parse_memory(c.resources.requests.get('memory', '0'))
+                            
+                key = f"{pod.metadata.namespace}/{pod.metadata.name}"
+                if key in pod_metrics:
+                    pod_cpu = sum([parse_cpu(c['usage'].get('cpu', '0')) for c in pod_metrics[key]])
+                    pod_mem = sum([parse_memory(c['usage'].get('memory', '0')) for c in pod_metrics[key]])
+                else:
+                    pod_cpu = pod_cpu_req
+                    pod_mem = pod_mem_req
                     
-                    if pod.spec.containers:
+                # D3 Treemap doesn't like 0 values, so minimum size 1
-                        for c in pod.spec.containers:
+                pod_cpu_val = max(pod_cpu, 1)
-                            if c.resources and c.resources.requests:
+                pod_mem_val = max(pod_mem, 1)
-                                pod_cpu_req += parse_cpu(c.resources.requests.get('cpu', '0'))
+                
-                                pod_mem_req += parse_memory(c.resources.requests.get('memory', '0'))
+                node_cpu_total += pod_cpu
-                                
+                node_mem_total += pod_mem
-                    key = f"{pod.metadata.namespace}/{pod.metadata.name}"
-                    if key in pod_metrics:
-                        pod_cpu = sum([parse_cpu(c['usage'].get('cpu', '0')) for c in pod_metrics[key]])
-                        pod_mem = sum([parse_memory(c['usage'].get('memory', '0')) for c in pod_metrics[key]])
-                    else:
-                        pod_cpu = pod_cpu_req
-                        pod_mem = pod_mem_req
-                        
-                    # D3 Treemap doesn't like 0 values, so minimum size 1
-                    pod_cpu_val = max(pod_cpu, 1)
-                    pod_mem_val = max(pod_mem, 1)
                     
-                    node_cpu_total += pod_cpu
+                node_data["children"].append({
-                    node_mem_total += pod_mem
+                    "name": pod.metadata.name,
-                        
+                    "namespace": pod.metadata.namespace,
-                    node_data["children"].append({
+                    "type": "pod",
-                        "name": pod.metadata.name,
+                    "cpu": pod_cpu_val,
-                        "namespace": pod.metadata.namespace,
+                    "memory": pod_mem_val,
-                        "type": "pod",
+                    "cpu_raw": pod_cpu,
-                        "cpu": pod_cpu_val,
+                    "memory_raw": pod_mem,
-                        "memory": pod_mem_val,
+                    "status": pod.status.phase
-                        "cpu_raw": pod_cpu,
+                })
-                        "memory_raw": pod_mem,
-                        "status": pod.status.phase
-                    })
             
             # Optionally add an "Unused" block so the node size matches allocatable
             unused_cpu = max(0, cpu_allocatable - node_cpu_total)
@@ -217,6 +266,10 @@ def cluster_data():
 def version_route():
     return jsonify({"version": get_version()})
 
+@app.route('/healthz')
+def healthz():
+    return jsonify({"status": "ok"}), 200
+
 @app.route('/')
 def index():
     return render_template('index.html', version=get_version())

helm-chart/Chart.yaml

@@ -2,5 +2,5 @@ apiVersion: v2
 name: k8s-pulse
 description: A Helm chart for k8s-pulse, a Kubernetes cluster resource visualizer.
 type: application
-version: 0.1.0
+version: 0.1.1
-appVersion: "1.0.4"
+appVersion: "1.0.8"

helm-chart/templates/deployment.yaml

@@ -42,13 +42,13 @@ spec:
               protocol: TCP
           livenessProbe:
             httpGet:
-              path: /
+              path: /healthz
               port: http
             initialDelaySeconds: 10
             periodSeconds: 10
           readinessProbe:
             httpGet:
-              path: /
+              path: /healthz
               port: http
             initialDelaySeconds: 5
             periodSeconds: 10

static/app.js

@@ -405,7 +405,8 @@ function build3DCluster(data) {
             isControlPlane: isControlPlane,
             roles: n.data.roles || (isControlPlane ? ['control-plane'] : ['worker']),
             internalIp: n.data.internalIp || n.data.internal_ip || 'N/A',
-            osImage: n.data.osImage || n.data.os_image || 'N/A'
+            osImage: n.data.osImage || n.data.os_image || 'N/A',
+            nonTerminatedPods: n.data.nonTerminatedPods ?? n.data.non_terminated_pods ?? n.data.podCount ?? n.data.pod_count ?? (n.data.children ? n.data.children.filter(c => c.type === 'pod' && c.status !== 'Succeeded' && c.status !== 'Failed').length : 0)
         };
         platformMesh.userData = nodeUserData;
 
@@ -601,6 +602,7 @@ function onPointerMove(event) {
             if (data.type === 'node') {
                 const isCP = data.isControlPlane || (data.roles && (data.roles.includes('control-plane') || data.roles.includes('master')));
                 const badgeLabel = data.roles && data.roles.length ? data.roles.join(', ') : (isCP ? 'control-plane' : 'Node');
+                const podCount = data.nonTerminatedPods ?? data.non_terminated_pods ?? data.podCount ?? data.pod_count ?? 0;
                 tooltip.innerHTML = `
                     <div class="tooltip-title">
                         <span>${data.name}</span>
@@ -614,6 +616,10 @@ function onPointerMove(event) {
                         <span class="tooltip-label">OS Image:</span>
                         <span class="tooltip-value">${data.osImage || 'N/A'}</span>
                     </div>
+                    <div class="tooltip-row">
+                        <span class="tooltip-label">Non-terminated Pods:</span>
+                        <span class="tooltip-value">${podCount}</span>
+                    </div>
                 `;
             } else if (data.type === 'unused') {
                 const valCpu = formatCpu(data.cpu_raw);

version.ini

@@ -1,2 +1,2 @@
 [metadata]
-version = 0.7
+version = 0.8