Grafana do not show any data under Kubernetes Cluster Details after upgrade TKGI from 1.13.x to 1.14.x.
search cancel

Grafana do not show any data under Kubernetes Cluster Details after upgrade TKGI from 1.13.x to 1.14.x.

book

Article ID: 298703

calendar_today

Updated On:

Products

VMware Tanzu Kubernetes Grid Integrated Edition

Issue/Introduction

Symptoms  
  • Customer has Ops Manager 2.10.52 / HealchWatch 2.2.4/ HealthWatch Exporter 2.2.4.
  • After customer upgrade TKGI from 1.13.x to 1.14.x, Grafana do not show cluster name list under “Kubernetes Cluster Details”
  • Other Dashboard works fine.


Environment

Product Version: 1.14

Resolution

Cause 
Started from TKGI 1.14 we get metrics from telegraf and we should change some config for the data visibility.
 

Resolution  
Apply below changes on your TKGI tile.
 
  1. Navigate to the Tanzu Kubernetes Grid Integrated Edition tile > Settings > Host Monitoring.
  2. Under Enable Telegraf Output, select Yes.
  3. Select “Include Kubernetes Scheduler metrics” and “Include telegraf metrics when telegraf enabled”
  4. Under “Setup Telegraf Output”, provide the following TOML configuration file. Please be noted, must input “metric_version = 2” value inside this file.
[[outputs.prometheus_client]]
listen = ":10200"
  metric_version = 2
  1. After edit "Setup Telegraf Outputs" >  apply-change> login to grafana UI > go to Kubernetes Cluster Detail > in the upper right corner, change time range to "last 15 minutes.
  2. After that, data will comes again. If above step cannot figure the issue, we need to check logs for pks-cluster-discovery job in tsdb vm under the healthwatch main tile because those metrics are fetched through PKS cluster discovery.