Customize Configurations of Monitoring Servers

When you deploy a TiDB cluster using TiUP, TiUP also deploys monitoring servers, such as Prometheus, Grafana, and Alertmanager. In the meantime, if you scale out this cluster, TiUP also adds the new nodes into monitoring scope.

To customize the configurations of the monitoring servers mentioned above, you can follow the instructions below to add related configuration items in the topology.yaml of the TiDB cluster.

Customize Prometheus configurations

Currently, TiUP supports customizing Prometheus rule and scrape configuration files.

Customize Prometheus rule configuration

1. Customize the rule configuration file and place it under a directory of the machine where TiUP locates.

2. In the topology.yaml file, set rule_dir to the directory of the customized rule configuration file.

   The following is a configuration example of monitoring_servers in the topology.yaml file:

   # # Server configs are used to specify the configuration of Prometheus Server.
   monitoring_servers:
     # # The ip address of the Monitoring Server.
   - host: 127.0.0.1
     rule_dir: /home/tidb/prometheus_rule   # prometheus rule dir on TiUP machine
   

After the preceding configuration is done, when you deploy, scale out, scale in, or reload a TiDB cluster, TiUP loads the customized rule configurations from rule_dir (for example, /home/tidb/prometheus_rule) and sends them to the Prometheus Server to replace the default rule configuration.

Customize Prometheus scrape configuration

1. Open the topology.yaml file of the TiDB cluster.

2. In the monitoring_servers configuration, add the additional_scrape_conf field.

   The following is a configuration example of monitoring_servers in the topology.yaml file:

   monitoring_servers:
   - host: xxxxxxx
   ssh_port: 22
   port: 9090
   deploy_dir: /tidb-deploy/prometheus-9090
   data_dir: /tidb-data/prometheus-9090
   log_dir: /tidb-deploy/prometheus-9090/log
   external_alertmanagers: []
   arch: amd64
   os: linux
   additional_scrape_conf:
     metric_relabel_configs:
       - source_labels: [__name__]
           separator: ;
           regex: tikv_thread_nonvoluntary_context_switches|tikv_thread_voluntary_context_switches|tikv_threads_io_bytes_total
           action: drop
       - source_labels: [__name__,name]
           separator: ;
           regex: tikv_thread_cpu_seconds_total;(tokio|rocksdb).+
           action: drop
   

After the preceding configuration is done, when you deploy, scale out, scale in, or reload a TiDB cluster, TiUP adds the additional_scrape_conf field to the corresponding parameters of the Prometheus configuration file.

Customize Grafana configurations

Currently, TiUP supports customizing Grafana Dashboard and other configurations.

Customize Grafana Dashboard

1. Customize the configuration file of the Grafana Dashboard and place it under a directory of the machine where TiUP locates.

2. In the topology.yaml file, set dashboard_dir to the directory of the customized Dashboard configuration file.

   The following is a configuration example of grafana_servers in the topology.yaml file:

   # # Server configs are used to specify the configuration of Grafana Servers.
   grafana_servers:
     # # The ip address of the Grafana Server.
    - host: 127.0.0.1
    dashboard_dir: /home/tidb/dashboards   # grafana dashboard dir on TiUP machine
   

After the preceding configuration is done, when you deploy, scale out, scale in, or reload a TiDB cluster, TiUP loads the customized Dashboard configurations from dashboard_dir (for example, /home/tidb/dashboards) and sends the configurations to the Grafana Server to replace the default Dashboard configuration.

Customize other Grafana configurations

1. Open the topology.yaml file of the TiDB cluster.

2. Add other configuration items in the grafana_servers configuration.

   The following is a configuration example of the [log.file] level and smtp fields in the topology.yaml file:

   # # Server configs are used to specify the configuration of Grafana Servers.
   grafana_servers:
   # # The ip address of the Grafana Server.
   - host: 127.0.0.1
       config:
       log.file.level: warning
       smtp.enabled: true
       smtp.host: {IP}:{port}
       smtp.user: example@pingcap.com
       smtp.password: {password}
       smtp.skip_verify: true
   

After the preceding configuration is done, when you deploy, scale out, scale in, or reload a TiDB cluster, TiUP adds the config field to the Grafana configuration file grafana.ini.

Customize Alertmanager configurations

Currently, TiUP supports customizing the listening address of Alertmanager.

Alertmanager deployed by TiUP listens to alertmanager_servers.host by default. You cannot access Alertmanager if you use a proxy. To address this issue, you can specify the listening address by adding listen_host to the cluster configuration file topology.yaml. The recommended value is 0.0.0.0.

The following example sets the listen_host field to 0.0.0.0.

alertmanager_servers:
  # # The ip address of the Alertmanager Server.
  - host: 172.16.7.147
    listen_host: 0.0.0.0
    # # SSH port of the server.
    ssh_port: 22

After the preceding configuration is done, when you deploy, scale out, scale in, or reload a TiDB cluster, TiUP adds the listen_host field to --web.listen-address in Alertmanager startup parameters.