filecoin-k8s

Repo for configuring and controlling the Filecoin Kubernetes cluster monitoring

Basic spinup for monitoring components

Requirements:

• helm v3.2+
• create slack webhook in your slack workspace, paste in Makefile SLACK-WEBHOOK-URL
• create or have public slack channel for alert messages, change it in Makefile SLACK-CHANNEL
• change nodeselector label in values.yaml, if your nodes don't have role: monitoring labels
• chage lotus node labels in values.yaml in additionalServiceMonitors section, if your lotus application doesn't use app: lotus-node-app label
• deploy ingress controller
• provide external dns name for grafana dashboard, change it in Makefile GRAFANA-HOST
• generate password for grafana dashboard login(login admin), change it in Makefile GRAFANA-WEB-PASSWORD

Default settings( may be changed in Makefile or in values.yaml ):

• helm chart name: monitoring
• monitoring namespace: monitoring(will be created and used)
• lotus node namespace: spacerace(you may change it to your own in file values.yaml in section prometheus.prometheusSpec.additionalServiceMonitors.namespaceSelector.matchNames)

Deploy monitoring

make create-monitoring

Upgrade monitoring

change parameters in values.yaml or customize Makefile with --set variable
make upgrade-monitoring

General usage

Once they are both spun up, you can login to grafana and choose necessary dashboard.

If you want to open prometheus, alertmanager ui locally , you may use kubectl proxy

Prometheus

http://127.0.0.1:8001/api/v1/namespaces/monitoring/services/prometheus-operated:9090/proxy

Alertmanager

http://127.0.0.1:8001/api/v1/namespaces/monitoring/services/alertmanager-operated:9093/proxy/