fix(monitoring): Prometheus 단일 복제본 대시보드 정합성

[Neibce]

Changelog

fix(monitoring): Prometheus 단일 복제본 대시보드 정합성 확보

[chatgpt-codex-connector]

💡 Codex Review

code-place/kubernetes/monitoring/prometheus-rules.yaml

Lines 65 to 66 in 54dc137

	- alert: PrometheusUnavailable
	expr: (sum(max by (pod) (kube_pod_status_ready{namespace="monitoring", condition="true", pod=~"prometheus-kube-prometheus-stack-prometheus-[0-9]+"})) < 1) or absent(kube_pod_status_ready{namespace="monitoring", condition="true", pod=~"prometheus-kube-prometheus-stack-prometheus-[0-9]+"})

Monitor the lone Prometheus from an external evaluator

When the single Prometheus pod or its node becomes unavailable, that same pod is no longer running the rule engine that evaluates this expression or sends the alert to Alertmanager. Consequently, the documented one-minute PrometheusUnavailable notification cannot fire for the primary outage scenario; retain another evaluator or use an external health/dead-man check.

---

code-place/kubernetes/base/postgres/postgres.yaml

Lines 20 to 21 in 54dc137

	storage:
	storageClass: longhorn-cnpg

Migrate existing CNPG volumes before switching classes

On either already-deployed dev or prod cluster, the bound PostgreSQL PVCs retain their immutable storageClassName: longhorn; changing the CNPG template here only affects newly provisioned claims and does not move the current database volumes onto the new single-replica Longhorn class. An explicit CNPG data/PVC migration or replacement procedure is needed, otherwise applying this change leaves the intended storage reduction unapplied to the existing databases.

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".