Kubernetes DaemonSet Guide: Deploy Pods to All Nodes

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87

# Kubernetes DaemonSet Guide: Deploy Pods to All Nodes

In this post, we dive deep into **Kubernetes DaemonSet** — a powerful controller that ensures a specific Pod runs on **every node** (or selected nodes) in your cluster.

We’ll explore:

* What a DaemonSet is and why it exists
* Common real-world use cases
* YAML examples and best practices
* Key differences from Deployment and StatefulSet
* Troubleshooting and common pitfalls

---

## Table of Contents

{% toc %}

---

## 1. What Is a DaemonSet?

A **DaemonSet** is a Kubernetes resource that ensures **one Pod runs on every node** in a cluster.

This is especially useful for cluster-wide background processes, like log collectors, system monitors, or GPU agents.

### Real-World Examples

* `fluentd` or `fluentbit` for log collection
* `prometheus-node-exporter` for metrics
* Custom agents for backups, GPU drivers, or node configuration

---

## 2. Why Use a DaemonSet?

If you want the **same Pod to run on all nodes**, a DaemonSet is the go-to solution.

| Use Case            | Description                                       |
| ------------------- | ------------------------------------------------- |
| System monitoring   | Node-level metrics collection (CPU, memory, etc.) |
| Logging             | Collect logs from all containers across nodes     |
| Node-specific setup | GPU drivers, firewall rules, or system agents     |
| DNS caching         | Cluster-wide DNS cache (e.g., `dnsmasq`)          |

Unlike a Deployment, which spreads replicas across the cluster, a DaemonSet guarantees **one Pod per node**, automatically.

---

## 3. Key Features of DaemonSet

| Feature                 | Description                                           |
| ----------------------- | ----------------------------------------------------- |
| One pod per node        | Ensures each node runs one instance of the Pod        |
| Auto-scaling with nodes | Adds/removes Pods as nodes join/leave the cluster     |
| Node targeting          | Use `nodeSelector` or `affinity` to control placement |
| Taint support           | With `tolerations`, you can run Pods on tainted nodes |

> Tip: DaemonSets are ideal for tasks that require full visibility across the cluster.

---

## 4. Basic DaemonSet YAML Example

```yaml
apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: log-agent
  namespace: kube-system
spec:
  selector:
    matchLabels:
      app: log-agent
  template:
    metadata:
      labels:
        app: log-agent
    spec:
      containers:
      - name: log-agent
        image: fluent/fluentd:latest
        resources:
          limits:
            memory: "200Mi"
            cpu: "200m"

1

image: prom/node-exporter

1
2
3

spec:
  nodeSelector:
    hardware: nvidia

1

kubectl edit daemonset log-agent -n kube-system

1

kubectl delete daemonset log-agent -n kube-system