viktor/infra
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

monitoring(prometheus): keep all 4 kubelet_volume_stats_inodes metrics

Browse source 
pvc-autoresizer's GetMetrics() returns volume stats for a PVC only if
all four kubelet_volume_stats metrics (available_bytes, capacity_bytes,
inodes_free, inodes) are retrieved. The keep-list in the
kubernetes-nodes scrape job had available_bytes and capacity_bytes
(post 9d5da4d8) but was missing the two inode metrics, so the
autoresizer's reconcile logged "failed to get volume stats" for every
PVC and never resized anything.

Add kubelet_volume_stats_inodes and kubelet_volume_stats_inodes_free
to the regex.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This commit is contained in:
Viktor Barzin 2026-05-10 18:36:16 +00:00
parent 6d3308c848
commit 4103ea2ba0
1 changed files with 6 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

10
stacks/monitoring/modules/monitoring/prometheus_chart_values.tpl
View file

@ -380,11 +380,13 @@ serverFiles:
regex: 'kubernetes_feature_enabled|kubelet_container_log_filesystem_used_bytes'
action: drop
# Whitelist: only keep essential kubelet metrics
# kubelet_volume_stats_available_bytes is required by pvc-autoresizer
# (it computes utilization as 1 - available/capacity). Without it the
# autoresizer is silent for every PVC in the cluster.
# pvc-autoresizer requires all four kubelet_volume_stats metrics
# (available_bytes, capacity_bytes, inodes_free, inodes); it skips
# any PVC missing even one. Without inodes_free + inodes the
# autoresizer's GetMetrics returns empty for every PVC and the
# reconcile is a no-op cluster-wide.
- source_labels: [__name__]
regex: 'kubelet_volume_stats_capacity_bytes|kubelet_volume_stats_used_bytes|kubelet_volume_stats_available_bytes|kubelet_volume_stats_inodes_used|kubelet_running_containers|kubelet_runtime_operations_errors_total|process_cpu_seconds_total|process_resident_memory_bytes|process_start_time_seconds|go_memstats_alloc_bytes|up'
regex: 'kubelet_volume_stats_capacity_bytes|kubelet_volume_stats_used_bytes|kubelet_volume_stats_available_bytes|kubelet_volume_stats_inodes|kubelet_volume_stats_inodes_used|kubelet_volume_stats_inodes_free|kubelet_running_containers|kubelet_runtime_operations_errors_total|process_cpu_seconds_total|process_resident_memory_bytes|process_start_time_seconds|go_memstats_alloc_bytes|up'
action: keep
- job_name: kubernetes-nodes-cadvisor
scheme: https