The CronJob ran every 5 min on a vanilla python:3.12-alpine image, doing `apk add git` + `pip install uptime-kuma-api` from scratch on every invocation. Caught at ~3.2 MB/s on k8s-node4's root LV, contributing to ~8 MB/s sustained on the pve-data thin pool (sdc) — ~804 GB written over the prior 18 h. Commented out the kubernetes_cron_job_v1.status_page_pusher resource (kept ns / SA / RBAC / ConfigMap intact for trivial revert). Re-enable once a custom image with git + uptime-kuma-api baked in is published so no per-run cold install happens. status.viktorbarzin.me stops updating until then. |
||
|---|---|---|
| .. | ||
| index.html | ||
| main.tf | ||
| terragrunt.hcl | ||