Stephan Lo
f797af114b
Configure markdownlint with rules aligned to technical documentation standards and Docsy theme conventions. Design Decisions: - Enable core quality rules (heading hierarchy, consistent list styles) - Allow inline HTML for Docsy shortcodes and components - Permit bare URLs (common in technical documentation) - Make code block language hints optional (pragmatic for existing content) - Set maximum 2 consecutive blank lines (balanced readability) - Enforce single trailing newline (POSIX standard) - Use asterisk for unordered lists (consistency) - Allow 2-space list indentation (Markdown standard) Auto-fixed Issues: - Converted dash lists to asterisk lists (568 fixes) - Removed trailing spaces (211 fixes) - Added missing trailing newlines (74 fixes) - Added blank lines around lists and headings (100+ fixes) Remaining Style Warnings (intentionally accepted): - MD029: List numbering variations in meeting notes (75 instances) - MD036: Bold text for section headers in ADRs (13 instances) - MD025: Multiple H1 in notes/brainstorming docs (10 instances) - MD032/MD022: Minor spacing variations (15 instances) Test Results: ✅ Hugo build: 227 pages generated successfully ✅ HTML validation: No errors ✅ Link checking: All links valid (except dev-only livereload) ✅ Markdown linting: Only non-critical style warnings remain The configuration balances strict quality checks with pragmatic flexibility for diverse content types (documentation, ADRs, meeting notes, tutorials).
2.2 KiB
| title | description |
|---|---|
| Kube-prometheus-stack | Kube-prometheus-stack contains Kubernetes manifests, Prometheus and Grafana, including preconfigured dashboards |
Kube-prometheus-stack Overview
Grafana is an open-source monitoring solution that enables viusalization of metrics and logs. Prometheus is an open-source monitoring and alerting system which collects metrics from services and allows the metrics to be shown in Grafana.
Implementation Details
The application ist started in edfbuilder/kind/stacks/core/kube-prometheus.yaml. The application has the sync option spec.syncPolicy.syncOptions ServerSideApply=true. This is necessary, since kube-prometheus-stack exceeds the size limit for secrets and without this option a sync attempt will fail and throw an exception. The Helm values file edfbuilder/kind/stacks/core/kube-prometheus/values.yaml contains configuration values: grafana.additionalDataSources contains Loki as a Grafana Data Source. grafana.ingress contains the Grafana ingress configuratione, like the host url (cnoe.localtest.me). grafana.sidecar.dashboards contains necessary configurations so additional user defined dashboards are loaded when Grafana is started. grafana.grafana.ini.server contains configuration details that are necessary, so the ingress points to the correct url.
Start
Once Grafana is running it is accessible under https://cnoe.localtest.me/grafana. Many preconfigured dashboards can be used by klicking the menu option Dashboards.
Adding your own dashboards
The application edfbuilder/kind/stacks/core/kube-prometheus.yaml is used to import new Loki dashboards. Examples for imported dashboards can be found in the folder edfbuilder/kind/stacks/core/kube-prometheus/dashboards.
It is possible to add your own dashboards: Dashboards must be in JSON format. To add your own dashboard create a new ConfigMap in YAML format using onw of the examples as a blueprint. The new dashboard in JSON format has to be added as the value for data.k8s-dashboard-[...].json like in the examples. (It is important to use a unique name for data.k8s-dashboard-[...].json for each dashboard.)
Currently preconfigured dashboards include several dahboards for Loki and a dashboard to showcase Nginx-Ingress metrics.