UMEE GRAFANA DASHBOARD

• Grafana version: v8.3.3 [ upgrade ! ]
• Requirement: node-exporter ( otherwise hardware data will not be available )
• Enable prometheus metrics in $HOME/.umee/config/config.toml

[instrumentation]
prometheus = true

• By default tendermint serve metrics at port 26660, can be adjusted with:

prometheus_listen_addr = ":26660"

• Allow metrics port in firewal, use tunnels, subnets or prefered method.
• Publicly expose Prometheus metrics is bad idea, don't do this !
• Check if metrics visible from Prometheus instance by:

curl -s <UMEE node IP here>:26660/metrics

• Point Prometheus to metrics end-point and restart, follow example, use your instance IP in case Prometheus is separated:

  - job_name: 'umee_tendermint'
    static_configs:
    - targets: ['127.0.0.1:26660']
      labels:
        alias: 'umee_tendermint'
        instance: umee_server_x

• add node_exporter:

  - job_name: 'umee_hardware'
    static_configs:
    - targets: ['127.0.0.1:12345']
      labels:
        alias: 'umee_server'
        instance: umee_server_x

• Import attached grafana dashboard json
• setup alerts, each alert have short description
• ask for help in Discord

AUTOSTAKE SERVICE ( used in test net, only works with "test" keyring )

Example: autostake.service

[Unit]
Description=AUTOSTAKE SERVICE
After=network-online.target

[Service]

Type=simple

User=umee
Group=umee

ExecStart=/path/to/autostake.sh

Restart=on-failure
RestartSec=60

KillMode=process

[Install]

WantedBy=multi-user.target