kubernetes-monitoring
kubernetes-monitoring copied to clipboard
dNationCloud
Monitor your Kubernetes infrastructure and applications at a glance using semaphore principle (green/orange/red)
The code contains exceptions for virt-launcher PODs, this should be improved as it forces some workarounds for use-cases e.g. when the external node-exporter is going to be used
[grafonnet-lib](https://github.com/grafana/grafonnet-lib) is deprecated. Test and validate if a replacement [grafonnet](https://github.com/grafana/grafonnet) suits our needs and use it if so.
Investigate and fix scenarios where L0 reports e.g. `warning` issues on L1, but L1 contains also some criticals. This could be connected with data sources that are used on L0...
Jeremy Stanley (fungi) pointed out: * Colors aren't universal. in china for example, red means something is good * Also red-green color blindness is fairly common, so labels can be...
 The choicebox seems to be a redundant feature as there is only one possible option. We could either get rid of the choicebox entirely or add the other datasources...
At L1 dashboard we use `avg` aggregation for overall utilization, but on L2 - `Disk per Node` we use `max`. It can be confusing. Either aggregation functions should be the...
- [ ] add steps to the release pipeline to generate dashboard in JSON form and rules in YAML format for monitoring plugin purpouses - use or modify makefile commads
see: https://prometheus.io/docs/prometheus/latest/configuration/alerting_rules/
Metadata
Owner
Metadata
Monitor your Kubernetes infrastructure and applications at a glance using semaphore principle (green/orange/red)