From 86ad271b237ad55f052267104a0dbaadea647391 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Cl=C3=A9ment=20D=C3=A9siles?=
 <1536672+cdesiles@users.noreply.github.com>
Date: Tue, 29 Sep 2026 12:25:56 +0200
Subject: [PATCH] fix(grafana): align NVMe temperature thresholds with 0-70C
 rating

Dashboard temperature panel now uses NVMe-specific thresholds (orange
60, red 70) instead of HDD ones; alert descriptions reference the
rated range.
---
 roles/grafana/README.md                     |  3 +--
 roles/grafana/files/alerting-rules.yaml     | 13 ++++++-------
 roles/grafana/files/smartctl-dashboard.json |  4 ++--
 3 files changed, 9 insertions(+), 11 deletions(-)

diff --git a/roles/grafana/README.md b/roles/grafana/README.md
index 5f13361..1626544 100644
--- a/roles/grafana/README.md
+++ b/roles/grafana/README.md
@@ -220,8 +220,7 @@ rise/fall semantics as capacity.
 | NVMe | ≥70°C | high | 10m |
 | NVMe | ≥75°C | critical | 5m |
 
-NVMe runs hotter: 40-60°C is normal, 60-70°C is warm (acceptable under
-sustained load), throttling typically starts at 70-80°C.
+NVMe tiers follow the 0-70°C operating spec (dashboard: orange ≥60, red ≥70°C).
 
 ### Capacity tiers
 
diff --git a/roles/grafana/files/alerting-rules.yaml b/roles/grafana/files/alerting-rules.yaml
index 1c775d9..df9a1ba 100644
--- a/roles/grafana/files/alerting-rules.yaml
+++ b/roles/grafana/files/alerting-rules.yaml
@@ -7,7 +7,7 @@ apiVersion: 1
 # Split by device class: HDDs (sd*) and NVMe (nvme*) have different
 # safe ranges, so they get separate tier ladders.
 #   HDD:  48..60°C (below 45 ideal) — device!~"(/dev/)?nvme.*"
-#   NVMe: 65..75°C (40-60 normal, 60-70 warm, >70 throttling) — device=~"(/dev/)?nvme.*"
+#   NVMe: 65..75°C (rated 0-70°C) — device=~"(/dev/)?nvme.*"
 # Capacity metrics: zfs_pool_capacity_percent and
 # node_filesystem_avail_bytes/node_filesystem_size_bytes (job="node").
 # node_filesystem_* is filtered to real local filesystems (ext4/xfs/btrfs);
@@ -411,9 +411,8 @@ groups:
           description: "S.M.A.R.T. temperature is below the 20°C ideal minimum (sensor/fan/offline disk?)."
 
       # --- NVMe temperature tiers (device=~"(/dev/)?nvme.*") ---
-      # NVMe runs hotter than HDD: 40-60°C is normal, 60-70°C is warm but
-      # acceptable under sustained load, and throttling typically starts in the
-      # 70-80°C range. The HDD ladder above excludes NVMe devices.
+      # Rated 0-70°C: <60 normal, 65 warm, 70 at the rated ceiling, 75 above
+      # spec. The HDD ladder above excludes NVMe devices.
       - uid: nvme_temp_ge_65
         title: "NVMe temperature ≥65°C"
         condition: C
@@ -460,7 +459,7 @@ groups:
           severity: warning
         annotations:
           summary: "NVMe {{ $labels.device }} at {{ $values.B }}°C ({{ $labels.instance }})"
-          description: "NVMe temperature is at or above 65°C (warm). Normal is 40-60°C; acceptable under sustained load."
+          description: "NVMe temperature is at or above 65°C (warm), approaching the 70°C rated maximum."
 
       - uid: nvme_temp_ge_70
         title: "NVMe temperature ≥70°C"
@@ -508,7 +507,7 @@ groups:
           severity: high
         annotations:
           summary: "NVMe {{ $labels.device }} at {{ $values.B }}°C ({{ $labels.instance }})"
-          description: "NVMe temperature is at or above 70°C. Thermal throttling typically begins in the 70-80°C range."
+          description: "NVMe temperature is at or above 70°C, the rated maximum operating temperature."
 
       - uid: nvme_temp_ge_75
         title: "NVMe temperature ≥75°C"
@@ -556,7 +555,7 @@ groups:
           severity: critical
         annotations:
           summary: "NVMe {{ $labels.device }} at {{ $values.B }}°C ({{ $labels.instance }})"
-          description: "NVMe temperature is at or above 75°C (hot). Check airflow/heatsink; sustained throttling expected."
+          description: "NVMe temperature is at or above 75°C, above the 70°C rated maximum. Check airflow/heatsink."
 
       # --- Capacity: ZFS pools (data disk) ---
       # zfs_pool_capacity_percent is emitted hourly by the zfs role's
diff --git a/roles/grafana/files/smartctl-dashboard.json b/roles/grafana/files/smartctl-dashboard.json
index 3c3ec53..d6a698e 100644
--- a/roles/grafana/files/smartctl-dashboard.json
+++ b/roles/grafana/files/smartctl-dashboard.json
@@ -54,9 +54,9 @@
       "title": "Temperature",
       "gridPos": { "h": 4, "w": 6, "x": 6, "y": 0 },
       "datasource": { "type": "prometheus", "uid": "prometheus" },
-      "fieldConfig": { "defaults": { "unit": "celsius", "color": { "mode": "thresholds" }, "thresholds": { "mode": "absolute", "steps": [ { "color": "green", "value": null }, { "color": "orange", "value": 45 }, { "color": "red", "value": 55 } ] } }, "overrides": [] },
+      "fieldConfig": { "defaults": { "unit": "celsius", "color": { "mode": "thresholds" }, "thresholds": { "mode": "absolute", "steps": [ { "color": "green", "value": null }, { "color": "orange", "value": 45 }, { "color": "red", "value": 55 } ] } }, "overrides": [ { "matcher": { "id": "byRegexp", "options": "^(/dev/)?nvme.*" }, "properties": [ { "id": "thresholds", "value": { "mode": "absolute", "steps": [ { "color": "green", "value": null }, { "color": "orange", "value": 60 }, { "color": "red", "value": 70 } ] } } ] } ] },
       "options": { "reduceOptions": { "values": false, "calcs": ["lastNotNull"], "fields": "" }, "colorMode": "background", "graphMode": "none", "textMode": "auto", "justifyMode": "auto", "orientation": "auto" },
-      "targets": [ { "refId": "A", "expr": "smartctl_device_temperature{instance=~\"$instance\", device=~\"$device\", temperature_type=\"current\"}" } ]
+      "targets": [ { "refId": "A", "expr": "smartctl_device_temperature{instance=~\"$instance\", device=~\"$device\", temperature_type=\"current\"}", "legendFormat": "{{device}}" } ]
     },
     {
       "id": 3,