參考示例demo圖

執行

git clone https://github.com/grafana/mimir.git

cd mimir

cd docs/sources/tutorials/play-with-grafana-mimir/

docker-compose up

參考配置

• prometheus 租戶支援

global:

  scrape_interval: 5s

  external_labels:

    cluster: demo

    namespace: demo

rule_files:
 

  - /etc/prometheus/rules.yaml

scrape_configs:

  - job_name: demo/mimir

    static_configs:

      - targets: [ 'mimir-1:8080' ]

        labels:

          pod: 'mimir-1'

      - targets: [ 'mimir-2:8080' ]

        labels:

          pod: 'mimir-2'

      - targets: [ 'mimir-3:8080' ]

        labels:
 

          pod: 'mimir-3'

remote_write:

  - url: http://load-balancer:9009/api/v1/push

# Add X-Scope-OrgID header so that Mimir knows what tenant the remote write data should be stored in.

# In this case, our tenant is "demo"

    // 基於header 請求頭

    headers:

      X-Scope-OrgID: demo

• mimir 配置
  比cortex簡單很多了，配置叢集，配置s3 儲存，配置規則，配置儲存

# Do not use this configuration in production.

# It is for demonstration purposes only.

# Run Mimir in single process mode, with all components running in 1 process. 

target: all,alertmanager,overrides-exporter

# Configure Mimir to use Minio as its object storage backend. 

blocks_storage:

  backend: s3

  s3:

    endpoint: minio:9000

    bucket_name: mimir-blocks

    access_key_id: mimir

    secret_access_key: supersecret

    insecure: true

  tsdb:

    dir: /data/ingester

# Use memberlist, a gossip-based protocol, to enable the 3 Mimir replicas to communicate 

memberlist:

  join_members: [ mimir-1, mimir-2, mimir-3 ]

ruler:

  rule_path: /data/ruler

  alertmanager_url: http://127.0.0.1:8080/alertmanager

  ring:

    # Quickly detect unhealthy rulers to speed up the tutorial.

    heartbeat_period: 2s

    heartbeat_timeout: 10s

ruler_storage:

  backend: s3

  s3:

    endpoint: minio:9000

    bucket_name: mimir-ruler

    access_key_id: mimir

    secret_access_key: supersecret

    insecure: true

alertmanager:

  data_dir: /data/alertmanager

  fallback_config_file: /etc/alertmanager-fallback-config.yaml

  external_url: http://localhost:9009/alertmanager

alertmanager_storage:

  backend: s3

  s3:

    endpoint: minio:9000

    bucket_name: mimir-alertmanager

    access_key_id: mimir

    secret_access_key: supersecret

    insecure: true

server:

  log_level: warn

效果

prometheus metrics

write metrics

說明

通過體驗mimir 是極大的簡化的metrics 的儲存處理，同時提供了多租戶的支援，後續還得深入研究下

參考資料

https://grafana.com/oss/mimir/
https://grafana.com/tutorials/play-with-grafana-mimir/?pg=oss-mimir&plcmt=hero-btn-1
https://grafana.com/blog/2022/03/30/announcing-grafana-mimir/?pg=oss-mimir&plcmt=hero-btn-3