About

Initial test example. Not everything is rendered or usable.

Tested with

HBase on k8s:

• k8s 1.15.x
• prometheus v2.17.2 via prometheus-operator
• grafana v6.7.3 (a04ef6cefc) with prometheus as Data Source
• Hadoop 3.1.3 (ninja deployment on kubernetes ;))
• hbase 2.2.4 with spdigital/prometheus-jmx-exporter-kubernetes:0.3.1 (default) via gradiant/charts

Known limitations

• You need to create service monitor on your own.
• You may need to adjust dashboard to match your prometheus labels
• Vast majority of metrics is not yet exposed on the graphs, but you can add them on your own
• Some metrics are duplicated because the way they are exposed by jmx_exporter

Changelog

• Fixed selector used for 'instances' - it was previously getting all instances existing for any metric, and now it only takes the instances which are specific for region or master server only. This dramatically speeds up graph loading

• Initial release

Contact

• contact author if you have any suggestions etc