clickhouse + prometheus + grafana 监控

prometheus + grafana 是已经存在的监控平台，此处不在赘述。只是介绍如何将 clickhouse 的监控数据接入 prometheus。

export 使用的是 clickhouse_exporter 。

具体步骤如下：

git clone https://github.com/ClickHouse/clickhouse_exporter  

# go build 需要添加下go的代理，此处采用阿里的 (需要预装go语言)
export GOPROXY=https://mirrors.aliyun.com/goproxy/
# 编译
go build 

# 信息配置
export CLICKHOUSE_USER="xx"
export CLICKHOUSE_PASSWORD="xxx"

# 后台启动
nohup ./clickhouse_exporter -scrape_uri=http://xxx.xxx.xxx.xxx:8123 > nohup.log 2>&1

# 验证是否成功
curl localhost:9116/metrics | head

 如果出现下面情况，证明已成功：

[root@ck02 clickhouse_exporter]# curl localhost:9116/metrics | head


  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 45573  100 45573   # HELP clickhouse_arena_alloc_bytes_total Number of ArenaAllocBytes total processed
 # TYPE clickhouse_arena_alloc_bytes_total counter
0clickhouse_arena_alloc_bytes_total 9.67172096e+08
 # HELP clickhouse_arena_alloc_chunks_total Number of ArenaAllocChunks total processed
 # TYPE clickhouse_arena_alloc_chunks_total counter
 clickhouse_arena_alloc_chunks_total 235666
 # HELP clickhouse_background_buffer_flush_schedule_pool_task Number of BackgroundBufferFlushSchedulePoolTask currently processed
 # TYPE clickhouse_background_buffer_flush_schedule_pool_task gauge
0clickhouse_background_buffer_flush_schedule_pool_task 0
 # HELP clickhouse_background_distributed_schedule_pool_task Number of BackgroundDistributedSchedulePoolTask currently processed
 1711k      0 --:--:-- --:--:-- --:--:-- 1780k

接下来修改下prometheus的配置：

scrape_configs:  
  # The job name is added as a label `job=<job_name>` to any timeseries scraped from this config.  
  - job_name: 'clickhouse'  

    # metrics_path defaults to '/metrics'  
    # scheme defaults to 'http'.  
    static_configs:  
    - targets: ['clickhouseexporter1:9116', 'clickhouseexporter2:9116', ...]

 找一个clickhouse的监控模板，这里我用了 882 模板。模板跳转。

最终效果图：

 刚接入，信息不多，整个流程就是这样子。