IoT edge clusters are implemented based on alarm notifications from Kubernetes Events

This article was last updated on: February 7, 2024 pm

background

Edge clusters (based on Raspberry Pi + K3S) need to implement basic alerting functions.

Edge cluster limits

  1. CPU/memory/storage resources are tight and cannot support a complete Prometheus-based monitoring system solution that requires at least 2GB of memory and a large amount of storage (even if it is based on Prometheus Agent) (Need to avoid additional storage and computing resource consumption)
  2. Network conditions cannot support the monitoring system, because the monitoring system generally needs to transmit data every 1min (or every moment), and the amount of data is not small;
    1. There is a 5G charging network, and the destination address needs to be activated and charged according to traffic, and due to 5G network conditions, the network transmission capacity is limited and unstable (it may be offline for a period of time);

Critical needs

To summarize, the key requirements are as follows:

  1. To implement timely alarms for edge cluster exceptions, you need to know the abnormal conditions that are occurring in the edge cluster.
  2. Network: The network conditions are poor, the network traffic is small, only a small number of end addresses can be opened, and the network is unstable (offline for a period of time) can be tolerated;
  3. Resources: You need to avoid additional storage and compute resource consumption as much as possible

scheme

In summary, the following scheme is adopted:

Alarm notifications based on Kubernetes Events

Architecture diagram

kubernetes-events-arch

Technical scenario planning

  1. Gather Events from various Kubernetes resources, such as:
    1. pod
    2. node
    3. kubelet
    4. crd
    5. …
  2. Pass kubernetes-event-exporter components to implement the collection of Kubernetes Events;
  3. Filter only Warning Level Events for alert notification (subsequently, conditions can be further defined)
  4. Alerts are sent through communication tools such as Feishu webhooks (subsequently, the sending channels can be increased)

Implementation steps

Manual way:

On the edge cluster, perform the following operations:

1. Create roles

kubectl apply The following yaml:

📝Notes:
Can be used cat << _EOF_ | kubectl apply -f - to create quickly

1
2
3
4
apiVersion: v1
kind: Namespace
metadata:
  name: monitoring
1
2
3
4
5
6
7
8
9
10
11
12
13
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: event-exporter-extra
rules:
  - apiGroups:
      - ""
    resources:
      - nodes
    verbs:
      - get
      - list
      - watch
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
apiVersion: v1
kind: ServiceAccount
metadata:
  namespace: monitoring
  name: event-exporter

```yaml
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: event-exporter
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: view
subjects:
  - kind: ServiceAccount
    namespace: monitoring
    name: event-exporter
1
2
3
4
5
6
7
8
9
10
11
12
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: event-exporter-extra
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: event-exporter-extra
subjects:
  - kind: ServiceAccount
    namespace: kube-event-export
    name: event-exporter

2. Create kubernetes-event-exporter config

kubectl apply The following yaml:

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
apiVersion: v1
kind: ConfigMap
metadata:
  name: event-exporter-cfg
  namespace: monitoring
data:
  config.yaml: |
    logLevel: error
    logFormat: json
    route:
      routes:
        - match:
            - receiver: "dump"      
        - drop:
            - type: "Normal"
          match:
            - receiver: "feishu"                     
    receivers:
      - name: "dump"
        stdout: {}
      - name: "feishu"
        webhook:
          endpoint: "https://open.feishu.cn/open-apis/bot/v2/hook/..."
          headers:
            Content-Type: application/json
          layout:
            msg_type: interactive
            card:
              config:
                wide_screen_mode: true
                enable_forward: true
              header:
                title: IoT edge clusters are implemented based on alarm notifications from Kubernetes Events
                  content: XXX IoT K3S 集群告警 
                template: red
              elements:
                - tag: div
                  text: 
                    tag: lark_md
                    content: "**EventType:**  {{ .Type }}\n**EventKind:**  {{ .InvolvedObject.Kind }}\n**EventReason:**  {{ .Reason }}\n**EventTime:**  {{ .LastTimestamp }}\n**EventMessage:**  {{ .Message }}"

🐾 Note:

  • endpoint: "https://open.feishu.cn/open-apis/bot/v2/hook/..." Change to the corresponding webhook endpoint as needed, ❌ remember not to announce it!!!
  • content: XXX IoT K3S 集群告警 : Adjust the name as needed for quick identification, such as: “Test K3S cluster alarm at home”

3. Create a deployment

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
cat << _EOF_ | kubectl apply -f -
apiVersion: apps/v1
kind: Deployment
metadata:
  name: event-exporter
  namespace: monitoring
spec:
  replicas: 1
  selector:
    matchLabels:
      app: event-exporter
      version: v1
  template:
    metadata:
      labels:
        app: event-exporter
        version: v1
    spec:
      volumes:
        - name: cfg
          configMap:
            name: event-exporter-cfg
            defaultMode: 420
        - name: localtime
          hostPath:
            path: /etc/localtime
            type: ''
        - name: zoneinfo
          hostPath:
            path: /usr/share/zoneinfo
            type: ''
      containers:
        - name: event-exporter
          image: ghcr.io/opsgenie/kubernetes-event-exporter:v0.11
          args:
            - '-conf=/data/config.yaml'
          env:
            - name: TZ
              value: Asia/Shanghai
          volumeMounts:
            - name: cfg
              mountPath: /data
            - name: localtime
              readOnly: true
              mountPath: /etc/localtime
            - name: zoneinfo
              readOnly: true
              mountPath: /usr/share/zoneinfo
          imagePullPolicy: IfNotPresent
      serviceAccount: event-exporter
      affinity:
        nodeAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
            - weight: 100
              preference:
                matchExpressions:
                  - key: node-role.kubernetes.io/controlplane
                    operator: In
                    values:
                      - 'true'
            - weight: 100
              preference:
                matchExpressions:
                  - key: node-role.kubernetes.io/control-plane
                    operator: In
                    values:
                      - 'true'
            - weight: 100
              preference:
                matchExpressions:
                  - key: node-role.kubernetes.io/master
                    operator: In
                    values:
                      - 'true'    
      tolerations:
        - key: node-role.kubernetes.io/controlplane
          value: 'true'
          effect: NoSchedule
        - key: node-role.kubernetes.io/control-plane
          operator: Exists
          effect: NoSchedule
        - key: node-role.kubernetes.io/master
          operator: Exists
          effect: NoSchedule      
_EOF_

📝 Illustrate:

  1. event-exporter-cfg The relevant configuration is used to load the configuration file saved in the form of ConfigMap;
  2. localtime zoneinfo TZ The relevant configuration is used to modify the time zone of the podAsia/Shanghaito make the final displayed notification effect in the CST time zone;
  3. affinity tolerations The relevant configuration is to ensure that: in any case, priority is scheduled to the master node and adjusted as needed, here because the master often exists as a gateway in the edge cluster, the configuration is high, and the online time is longer;

Automate deployment

Effect:Deploy automatically when K3S is installed

On the node where the K3S server is located,/var/lib/rancher/k3s/server/manifests/ Directory (created first if it doesn’t exist), Create event-exporter.yaml

1
2
3
4
apiVersion: v1
kind: Namespace
metadata:
  name: monitoring
1
2
3
4
5
6
7
8
9
10
11
12
13
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: event-exporter-extra
rules:
  - apiGroups:
      - ""
    resources:
      - nodes
    verbs:
      - get
      - list
      - watch
1
2
3
4
5
apiVersion: v1
kind: ServiceAccount
metadata:
  namespace: monitoring
  name: event-exporter
1
2
3
4
5
6
7
8
9
10
11
12
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: event-exporter
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: view
subjects:
  - kind: ServiceAccount
    namespace: monitoring
    name: event-exporter
1
2
3
4
5
6
7
8
9
10
11
12
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: event-exporter-extra
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: event-exporter-extra
subjects:
  - kind: ServiceAccount
    namespace: kube-event-export
    name: event-exporter
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
apiVersion: v1
kind: ConfigMap
metadata:
  name: event-exporter-cfg
  namespace: monitoring
data:
  config.yaml: |
    logLevel: error
    logFormat: json
    route:
      routes:
        - match:
            - receiver: "dump"      
        - drop:
            - type: "Normal"
          match:
            - receiver: "feishu"                     
    receivers:
      - name: "dump"
        stdout: {}
      - name: "feishu"
        webhook:
          endpoint: "https://open.feishu.cn/open-apis/bot/v2/hook/dc4fd384-996b-4d20-87cf-45b3518869ec"
          headers:
            Content-Type: application/json
          layout:
            msg_type: interactive
            card:
              config:
                wide_screen_mode: true
                enable_forward: true
              header:
                title: IoT edge clusters are implemented based on alarm notifications from Kubernetes Events
                  content: xxxK3S 集群告警 
                template: red
              elements:
                - tag: div
                  text: 
                    tag: lark_md
                    content: "**EventType:**  {{ .Type }}\n**EventKind:**  {{ .InvolvedObject.Kind }}\n**EventReason:**  {{ .Reason }}\n**EventTime:**  {{ .LastTimestamp }}\n**EventMessage:**  {{ .Message }}"
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
apiVersion: apps/v1
kind: Deployment
metadata:
  name: event-exporter
  namespace: monitoring
spec:
  replicas: 1
  selector:
    matchLabels:
      app: event-exporter
      version: v1
  template:
    metadata:
      labels:
        app: event-exporter
        version: v1
    spec:
      volumes:
        - name: cfg
          configMap:
            name: event-exporter-cfg
            defaultMode: 420
        - name: localtime
          hostPath:
            path: /etc/localtime
            type: ''
        - name: zoneinfo
          hostPath:
            path: /usr/share/zoneinfo
            type: ''
      containers:
        - name: event-exporter
          image: ghcr.io/opsgenie/kubernetes-event-exporter:v0.11
          args:
            - '-conf=/data/config.yaml'
          env:
            - name: TZ
              value: Asia/Shanghai
          volumeMounts:
            - name: cfg
              mountPath: /data
            - name: localtime
              readOnly: true
              mountPath: /etc/localtime
            - name: zoneinfo
              readOnly: true
              mountPath: /usr/share/zoneinfo
          imagePullPolicy: IfNotPresent
      serviceAccount: event-exporter
      affinity:
        nodeAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
            - weight: 100
              preference:
                matchExpressions:
                  - key: node-role.kubernetes.io/controlplane
                    operator: In
                    values:
                      - 'true'
            - weight: 100
              preference:
                matchExpressions:
                  - key: node-role.kubernetes.io/control-plane
                    operator: In
                    values:
                      - 'true'
            - weight: 100
              preference:
                matchExpressions:
                  - key: node-role.kubernetes.io/master
                    operator: In
                    values:
                      - 'true'    
      tolerations:
        - key: node-role.kubernetes.io/controlplane
          value: 'true'
          effect: NoSchedule
        - key: node-role.kubernetes.io/control-plane
          operator: Exists
          effect: NoSchedule
        - key: node-role.kubernetes.io/master
          operator: Exists
          effect: NoSchedule

After that, starting K3S will be deployed automatically.

📚️Reference:
Automatic deployment of manifests and Helm charts Rancher documentation

The final effect

As shown in the following figure:

image-20220413122040530

📚️ Reference documentation

Observability
#K8S #CloudNative #Observability #Prometheus #K3S #IoT #Edge
IoT edge clusters are implemented based on alarm notifications from Kubernetes Events
https://e-whisper.com/posts/48832/
Author
east4ming
Posted on
April 13, 2022
Licensed under