我准备怎么用一台服务器做完一次运维实践任务：从环境准备到监控与日志的落地路线

1
- hosts: k3s_nodes
2
  become: true
3
  tasks:
4
    - name: Update apt cache
5
      apt:
6
        update_cache: true
7

8
    - name: Install common packages
9
      apt:
10
        name:
11
          - curl
12
          - wget
13
          - git
14
          - vim
15
          - htop
16
          - unzip
17
          - jq
18
          - ca-certificates
19
        state: present

1
- hosts: k3s_nodes
2
  become: true
3
  tasks:
4
    - name: Install k3s
5
      shell: curl -sfL https://get.k3s.io | sh -
6
      args:
7
        creates: /usr/local/bin/k3s
8

9
    - name: Ensure k3s service is enabled
10
      service:
11
        name: k3s
12
        state: started
13
        enabled: true
14

15
    - name: Create kubectl symlink
16
      file:
17
        src: /usr/local/bin/k3s
18
        dest: /usr/local/bin/kubectl
19
        state: link
20
        force: true
21

22
    - name: Check node status
23
      shell: kubectl get nodes
24
      environment:
25
        KUBECONFIG: /etc/rancher/k3s/k3s.yaml
26
      register: k3s_nodes_result
27
      changed_when: false
28

29
    - name: Show node status
30
      debug:
31
        var: k3s_nodes_result.stdout_lines
32

33
    - name: Install Helm
34
      shell: curl https://raw.githubusercontent.com/helm/helm/main/scripts/get-helm-3 | bash
35
      args:
36
        creates: /usr/local/bin/helm

1
- hosts: k3s_nodes
2
  become: true
3
  environment:
4
    KUBECONFIG: /etc/rancher/k3s/k3s.yaml
5
    HTTP_PROXY: http://192.168.3.14:7890/
6
    HTTPS_PROXY: http://192.168.3.14:7890/
7
    NO_PROXY: 127.0.0.1,localhost,10.42.0.0/16,10.43.0.0/16,10.0.0.0/8,192.168.0.0/16
8
  tasks:
9
    - name: Add Prometheus Helm repo
10
      shell: helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
11
      register: helm_repo_add
12
      failed_when: helm_repo_add.rc != 0 and 'already exists' not in helm_repo_add.stderr
13
      changed_when: "'has been added' in helm_repo_add.stdout"
14

15
    - name: Add Grafana Helm repo
16
      shell: helm repo add grafana https://grafana.github.io/helm-charts
17
      register: grafana_repo_add
18
      failed_when: grafana_repo_add.rc != 0 and 'already exists' not in grafana_repo_add.stderr
19
      changed_when: "'has been added' in grafana_repo_add.stdout"
20

21
    - name: Update Helm repos
22
      shell: helm repo update
23
      changed_when: false
24

25
    - name: Create monitoring namespace
26
      shell: kubectl create namespace monitoring
27
      register: monitoring_ns
28
      failed_when: monitoring_ns.rc != 0 and 'already exists' not in monitoring_ns.stderr
29
      changed_when: "'created' in monitoring_ns.stdout"
30

31
    - name: Write monitoring values
32
      copy:
33
        dest: /root/monitoring-values.yaml
34
        mode: '0644'
35
        content: |
36
          grafana:
37
            service:
38
              type: NodePort
39
              nodePort: 30080
40

41
    - name: Install kube-prometheus-stack
42
      shell: helm upgrade --install monitoring prometheus-community/kube-prometheus-stack -n monitoring --create-namespace -f /root/monitoring-values.yaml

1
http://127.0.0.1:3000

1
- hosts: k3s_nodes
2
  become: true
3
  environment:
4
    KUBECONFIG: /etc/rancher/k3s/k3s.yaml
5
    HTTP_PROXY: http://192.168.3.14:7890/
6
    HTTPS_PROXY: http://192.168.3.14:7890/
7
    NO_PROXY: 127.0.0.1,localhost,10.42.0.0/16,10.43.0.0/16,10.0.0.0/8,192.168.0.0/16
8
  tasks:
9
    - name: Add Grafana Helm repo
10
      shell: helm repo add grafana https://grafana.github.io/helm-charts
11
      register: grafana_repo_add
12
      failed_when: grafana_repo_add.rc != 0 and 'already exists' not in grafana_repo_add.stderr
13
      changed_when: "'has been added' in grafana_repo_add.stdout"
14

15
    - name: Update Helm repos
16
      shell: helm repo update
17
      changed_when: false
18

19
    - name: Create logging namespace
20
      shell: kubectl create namespace logging
21
      register: logging_ns
22
      failed_when: logging_ns.rc != 0 and 'already exists' not in logging_ns.stderr
23
      changed_when: "'created' in logging_ns.stdout"
24

25
    - name: Write Loki values
26
      copy:
27
        dest: /root/loki-values.yaml
28
        mode: '0644'
29
        content: |
30
          deploymentMode: SingleBinary
31

32
          singleBinary:
33
            replicas: 1
34

35
          loki:
36
            auth_enabled: false
37
            commonConfig:
38
              replication_factor: 1
39
            storage:
40
              type: filesystem
41
            schemaConfig:
42
              configs:
43
                - from: "2024-01-01"
44
                  store: tsdb
45
                  object_store: filesystem
46
                  schema: v13
47
                  index:
48
                    prefix: loki_index_
49
                    period: 24h
50
            limits_config:
51
              allow_structured_metadata: true
52
              volume_enabled: true
53

54
          chunksCache:
55
            enabled: false
56
          resultsCache:
57
            enabled: false
58

59
          backend:
60
            replicas: 0
61
          read:
62
            replicas: 0
63
          write:
64
            replicas: 0
65
          ingester:
66
            replicas: 0
67
          querier:
68
            replicas: 0
69
          queryFrontend:
70
            replicas: 0
71
          queryScheduler:
72
            replicas: 0
73
          indexGateway:
74
            replicas: 0
75
          compactor:
76
            replicas: 0
77
          ruler:
78
            replicas: 0
79

80
          minio:
81
            enabled: false
82

83
          gateway:
84
            enabled: false
85

86
    - name: Install Loki
87
      shell: helm upgrade --install loki grafana/loki -n logging --create-namespace -f /root/loki-values.yaml
88

89
    - name: Write Alloy values
90
      copy:
91
        dest: /root/alloy-values.yaml
92
        mode: '0644'
93
        content: |
94
          alloy:
95
            configMap:
96
              create: true
97
              content: |-
98
                logging {
99
                  level  = "info"
100
                  format = "logfmt"
101
                }
102

103
                discovery.kubernetes "pods" {
104
                  role = "pod"
105
                }
106

107
                discovery.relabel "pod_logs" {
108
                  targets = discovery.kubernetes.pods.targets
109

110
                  rule {
111
                    source_labels = ["__meta_kubernetes_namespace"]
112
                    target_label  = "namespace"
113
                  }
114

115
                  rule {
116
                    source_labels = ["__meta_kubernetes_pod_name"]
117
                    target_label  = "pod"
118
                  }
119

120
                  rule {
121
                    source_labels = ["__meta_kubernetes_pod_container_name"]
122
                    target_label  = "container"
123
                  }
124
                }
125

126
                loki.source.kubernetes "pod_logs" {
127
                  targets    = discovery.relabel.pod_logs.output
128
                  forward_to = [loki.write.default.receiver]
129
                }
130

131
                loki.write "default" {
132
                  endpoint {
133
                    url = "http://loki.logging.svc.cluster.local:3100/loki/api/v1/push"
134
                  }
135
                }
136

137
          controller:
138
            type: daemonset
139

140
          serviceMonitor:
141
            enabled: false
142

143
    - name: Install Alloy
144
      shell: helm upgrade --install alloy grafana/alloy -n logging -f /root/alloy-values.yaml