diff --git a/.gitignore b/.gitignore
index 5fbc16d..e24ce90 100644
--- a/.gitignore
+++ b/.gitignore
@@ -190,3 +190,4 @@ pyrightconfig.json
# End of https://www.toptal.com/developers/gitignore/api/python,linux
myvalues.yaml
+
diff --git a/app.py b/app.py
index b454d7e..355e70b 100644
--- a/app.py
+++ b/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
- pods = v1.list_pod_for_all_namespaces().items
-
- node_metrics = {}
- pod_metrics = {}
- try:
- nm = custom.list_cluster_custom_object("metrics.k8s.io", "v1beta1", "nodes")
- for item in nm.get('items', []):
- node_metrics[item['metadata']['name']] = item['usage']
-
- pm = custom.list_cluster_custom_object("metrics.k8s.io", "v1beta1", "pods")
- for item in pm.get('items', []):
- key = f"{item['metadata']['namespace']}/{item['metadata']['name']}"
- pod_metrics[key] = item['containers']
- except ApiException:
- pass # Metrics server not available
+ v1, custom = get_k8s_client()
+
+ with ThreadPoolExecutor(max_workers=4) as executor:
+ fut_nodes = executor.submit(fetch_nodes, v1)
+ fut_pods = executor.submit(fetch_pods, v1)
+ fut_node_metrics = executor.submit(fetch_node_metrics, custom)
+ fut_pod_metrics = executor.submit(fetch_pod_metrics, custom)
+
+ nodes = fut_nodes.result()
+ pods = fut_pods.result()
+ node_metrics = fut_node_metrics.result()
+ pod_metrics = fut_pod_metrics.result()
+
+ cluster = {
+ "name": "Cluster",
+ "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:
- if pod.spec.node_name == node_name:
- pod_cpu_req = 0
- pod_mem_req = 0
+ for pod in node_pods:
+ pod_cpu_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:
- 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
-
- # 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)
+ # 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_mem_total += pod_mem
- node_cpu_total += pod_cpu
- node_mem_total += pod_mem
-
- node_data["children"].append({
- "name": pod.metadata.name,
- "namespace": pod.metadata.namespace,
- "type": "pod",
- "cpu": pod_cpu_val,
- "memory": pod_mem_val,
- "cpu_raw": pod_cpu,
- "memory_raw": pod_mem,
- "status": pod.status.phase
- })
+ node_data["children"].append({
+ "name": pod.metadata.name,
+ "namespace": pod.metadata.namespace,
+ "type": "pod",
+ "cpu": pod_cpu_val,
+ "memory": pod_mem_val,
+ "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())
diff --git a/helm-chart/Chart.yaml b/helm-chart/Chart.yaml
index 7f6ec69..a8c5111 100644
--- a/helm-chart/Chart.yaml
+++ b/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
-appVersion: "1.0.4"
+version: 0.1.1
+appVersion: "1.0.8"
diff --git a/helm-chart/templates/deployment.yaml b/helm-chart/templates/deployment.yaml
index 3793e4e..8829eb4 100644
--- a/helm-chart/templates/deployment.yaml
+++ b/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
diff --git a/static/2026-07-herdrbg.png b/static/2026-07-herdrbg.png
new file mode 100644
index 0000000..f807506
Binary files /dev/null and b/static/2026-07-herdrbg.png differ
diff --git a/static/app.js b/static/app.js
index f6fc70b..82357b0 100644
--- a/static/app.js
+++ b/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 = `
${data.name}
@@ -614,6 +616,10 @@ function onPointerMove(event) {
OS Image:
${data.osImage || 'N/A'}
+
+ Non-terminated Pods:
+ ${podCount}
+
`;
} else if (data.type === 'unused') {
const valCpu = formatCpu(data.cpu_raw);
diff --git a/version.ini b/version.ini
index 6f4c939..f501a3b 100644
--- a/version.ini
+++ b/version.ini
@@ -1,2 +1,2 @@
[metadata]
-version = 0.7
+version = 0.8