Exclude an external shelf by explicit config, not by path shape

Testing on a second host showed the internal/external split cannot be
inferred: iz-pve1's MD1200 sits behind its own HBA while iz-pve0's
internal drives sit behind a SAS expander, so the previous
"pci-*-scsi-*" glob matched every drive on one host and none on the
other. DISK_EXCLUDE_PATTERN now names the shelf's HBA per host.

Also fixes parse_disk_limit, which used awk's for-in over split fields
and so returned 0 or 70 from "0/70" depending on iteration order.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

README.md

@@ -127,6 +127,7 @@ systemctl enable --now fan_speed.service
 | `HDD_LIMIT_CAP` / `SSD_LIMIT_CAP` | `60` / `70` | Ceiling on what a drive may claim |
 | `TREND_SAMPLES` | `90` | Window length, in passes |
 | `TREND_RISE_ALARM` | `8` | Degrees of climb that alarms |
+| `DISK_EXCLUDE_PATTERN` | *(empty)* | by-path substring to skip, e.g. an external shelf's HBA |
 | `ALERT_EMAIL` | `Servers@ntfy1.izebra.xyz` | |
 | `ALERT_COOLDOWN` | `3600` | Seconds between repeats of one alarm |
 | `SMART_CHECK_INTERVAL` | `3600` | Seconds between SMART sweeps |
@@ -145,8 +146,12 @@ trusting them anywhere else.
 - Kernel device names are not stable — a shelf rescan renamed `sdaa`–`sdai` to
   `sds`–`sdaa` mid-session. Everything resolves through `/dev/disk/by-path` on every
   pass; never persist an `sdX`.
-- `DISK_GLOB` matches internal drives only (`pci-*-scsi-*`). Drives behind a SAS expander
+- **An external shelf must be excluded by hand.** Set `DISK_EXCLUDE_PATTERN` to its HBA's
-  are a separate enclosure with separate cooling and are deliberately excluded.
+  PCI address in the systemd unit. This cannot be inferred: `iz-pve1`'s MD1200 sits behind
+  its own HBA at `pci-0000:04:00.0`, while `iz-pve0`'s *internal* drives sit behind a SAS
+  expander at `pci-0000:02:00.0` — so "behind an expander" identifies an external
+  enclosure on one host and the internal backplane on the other. `iz-pve1` sets it,
+  `iz-pve0` has no shelf and leaves it empty.
 - Drive temperature limits are read once at startup. They do not change, and `smartctl -x`
   is far heavier than the `-A` used on the fast loop.
 - `smartctl -n standby` throughout, so a sleeping drive is skipped rather than spun up

fan_speed.service

@@ -2,7 +2,14 @@
 Description=Fan_Speed service
 
 [Service]
-ExecStart=/bin/bash -c /root/fan_speed/fan_speed.sh
+# Hosts with an external disk shelf must exclude that HBA - the shelf has its
+# own cooling and must not drive server fans. Find the PCI address with
+# `ls /dev/disk/by-path/`; there is no way to infer it (see monitor.sh).
+# iz-pve1 needs this; iz-pve0 has no shelf and needs nothing.
+#Environment=DISK_EXCLUDE_PATTERN=pci-0000:04:00.0
+ExecStart=/root/fan_speed/fan_speed.sh
+Restart=always
+RestartSec=10
 
 [Install]
 WantedBy=multi-user.target

monitor.sh

@@ -8,9 +8,17 @@
 
 # ---------------------------------------------------------------- configuration
 
-# Internal drives only. The DAS shelf enumerates behind a SAS expander
+DISK_GLOB=${DISK_GLOB:-/dev/disk/by-path/*}
-# (*-sas-exp*) and has its own controller - it must not drive server fans.
+
-DISK_GLOB=${DISK_GLOB:-/dev/disk/by-path/pci-*-scsi-*}
+# Drives to leave alone, matched against the by-path link. Set this to the PCI
+# address of an external shelf's HBA on any host that has one.
+#
+# There is no reliable way to infer this. An external enclosure looks exactly
+# like an internal backplane from /dev/disk/by-path: iz-pve1's MD1200 sits
+# behind its own HBA at pci-0000:04:00.0, while iz-pve0's *internal* drives sit
+# behind a SAS expander at pci-0000:02:00.0. Guessing from "-sas-exp" picks up
+# the wrong set on one host or the other, so it has to be stated per host.
+DISK_EXCLUDE_PATTERN=${DISK_EXCLUDE_PATTERN:-}
 
 # Each drive's fan ramp is derived from its own maximum operating temperature:
 #   ramp starts at  limit - DISK_RAMP_LOW_OFFSET
@@ -53,6 +61,9 @@ disk_devices() {
   local link dev
   for link in $DISK_GLOB; do
     case "$link" in *-part*) continue ;; esac
+    if [ -n "$DISK_EXCLUDE_PATTERN" ]; then
+      case "$link" in *"$DISK_EXCLUDE_PATTERN"*) continue ;; esac
+    fi
     [ -e "$link" ] || continue
     dev=$(readlink -f "$link")
     [ -b "$dev" ] && echo "$dev"
@@ -73,7 +84,10 @@ parse_disk_temperature() {
 # smartctl -x output on stdin -> the drive's own maximum operating temperature.
 parse_disk_limit() {
   awk '
-    /Min\/Max Temperature Limit/ { split($0, a, /[ \/]+/); for (i in a) if (a[i] ~ /^[0-9]+$/) v = a[i]; print v; exit }
+    /Min\/Max Temperature Limit/ {
+      if (match($0, /-?[0-9]+\/[0-9]+/)) { split(substr($0, RSTART, RLENGTH), a, "/"); print a[2] }
+      exit
+    }
     /Drive Trip Temperature/ { print $4; exit }'
 }