This dashboard allows PKS Operators to manage master nodes by examaning etcd metrics and infrastucture metrics from the node exporter. To configure PKS to export metrics to this dashboard input the following in the monitoring tab of the PKS tile.
[[outputs.wavefront]]
token="<your-direct-ingest-token>"
url="https://<your-instance>.wavefront.com"
[[processors.override]]
[processors.override.tags]
director = "<your-director-name>"
The following are some ideas around monitoring master nodes.
How many clusters do you have and what are they called? The dashboard takes
advantage of the cluster_name
tag and etcd.server.guage.id
metric.
Another concern for PKS operators is leadership election on Master
These "pressure" charts invert outbound network traffic and disk writes to help operators visualize the shape of key usage.
TBD
TBD