manatee: Update monitoring in home-assistant

- Add all disks to smartd
- Generate home-assistant config in nix
- Add metrics for all HDDs
This commit is contained in:
Alexander Heldt
2026-05-30 16:54:40 +00:00
parent d291633fe2
commit 4a63c4eb5e
4 changed files with 353 additions and 141 deletions
+5
View File
@@ -79,11 +79,16 @@
{ device = config.disko.devices.disk.root.device; }
{ device = config.disko.devices.disk.disk1.device; }
{ device = config.disko.devices.disk.disk2.device; }
{ device = config.disko.devices.disk.disk3.device; }
{ device = config.disko.devices.disk.disk4.device; }
];
};
services.zfs.autoScrub.enable = true;
# Don't force-import the pool if it appears in use elsewhere; safer default in 26.11+.
boot.zfs.forceImportRoot = false;
networking.hostId = "0a9474e7"; # Required by ZFS
disko.devices = {
disk = {
+1
View File
@@ -23,6 +23,7 @@ in
komga.enable = true;
romm.enable = true;
homepage.enable = true;
disk-smart.enable = true;
};
};
}
@@ -0,0 +1,159 @@
{
pkgs,
lib,
config,
...
}:
let
enabled = config.mod.disk-smart.enable;
disks = [
{ path = "/dev/disk/by-id/ata-ST8000VN004-3CP101_WWZ8QCG4"; name = "seagate_8tb_1"; label = "Seagate 8TB #1"; }
{ path = "/dev/disk/by-id/ata-ST8000VN004-3CP101_WWZ8QDJ5"; name = "seagate_8tb_2"; label = "Seagate 8TB #2"; }
{ path = "/dev/disk/by-id/ata-TOSHIBA_MG10ACA20TE_85K2A0UCF4MJ"; name = "toshiba_20tb_1"; label = "Toshiba 20TB #1"; }
{ path = "/dev/disk/by-id/ata-TOSHIBA_MG10ACA20TE_85K2A0V6F4MJ"; name = "toshiba_20tb_2"; label = "Toshiba 20TB #2"; }
];
outputDir = "/var/lib/disk-smart";
collectScript = pkgs.writeShellScript "disk-smart-collect" ''
set -euo pipefail
export PATH="${lib.makeBinPath [ pkgs.smartmontools pkgs.jq pkgs.coreutils ]}"
mkdir -p ${outputDir}
result="{"
${lib.concatMapStringsSep "\n" (disk: ''
raw=$(smartctl -j -A -H ${disk.path} 2>/dev/null || true)
temp=$(echo "$raw" | jq -r '.temperature.current // empty')
power_on=$(echo "$raw" | jq -r '.power_on_time.hours // empty')
smart_status=$(echo "$raw" | jq -r '.smart_status.passed // empty')
reallocated=$(echo "$raw" | jq -r '[.ata_smart_attributes.table[] | select(.name == "Reallocated_Sector_Ct")][0].raw.value // empty')
pending=$(echo "$raw" | jq -r '[.ata_smart_attributes.table[] | select(.name == "Current_Pending_Sector")][0].raw.value // empty')
result="$result\"${disk.name}\":{\"temperature\":$temp,\"power_on_hours\":$power_on,\"smart_passed\":$smart_status,\"reallocated_sectors\":$reallocated,\"pending_sectors\":$pending},"
'') disks}
# Remove trailing comma, close object
result="''${result%,}}"
echo "$result" | jq . > ${outputDir}/smart.json.tmp
mv ${outputDir}/smart.json.tmp ${outputDir}/smart.json
'';
indent = prefix: s:
lib.concatMapStringsSep "\n"
(line: if line == "" then line else prefix + line)
(lib.splitString "\n" s);
mkSensor = disk: ''
- name: "${disk.label} Temperature"
value_template: "{{ value_json.${disk.name}.temperature }}"
unit_of_measurement: "°C"
device_class: temperature
state_class: measurement
- name: "${disk.label} Power On Hours"
value_template: "{{ value_json.${disk.name}.power_on_hours }}"
unit_of_measurement: "h"
state_class: total_increasing
- name: "${disk.label} SMART Passed"
value_template: "{{ value_json.${disk.name}.smart_passed }}"
- name: "${disk.label} Reallocated Sectors"
value_template: "{{ value_json.${disk.name}.reallocated_sectors }}"
state_class: measurement
- name: "${disk.label} Pending Sectors"
value_template: "{{ value_json.${disk.name}.pending_sectors }}"
state_class: measurement
'';
sensorYaml = indent " " (lib.concatMapStrings mkSensor disks);
sectorEntities = lib.concatMap (disk: [
"sensor.${disk.name}_reallocated_sectors"
"sensor.${disk.name}_pending_sectors"
]) disks;
sectorEntitiesYaml = lib.concatMapStringsSep "\n"
(id: " - ${id}") sectorEntities;
smartPassedEntities = map (disk: "sensor.${disk.name}_smart_passed") disks;
smartPassedEntitiesYaml = lib.concatMapStringsSep "\n"
(id: " - ${id}") smartPassedEntities;
in
{
options = {
mod.disk-smart = {
enable = lib.mkEnableOption "Enable disk SMART monitoring module";
};
};
config = lib.mkIf enabled {
mod.home-assistant.extraConfig = ''
rest:
- resource: http://127.0.0.1:9633/smart.json
scan_interval: 60
sensor:
${sensorYaml}
automation disk_smart:
- alias: "Disk sector count increased"
trigger:
- platform: state
entity_id:
${sectorEntitiesYaml}
condition:
- condition: template
value_template: "{{ trigger.from_state.state | int(-1) >= 0 and trigger.to_state.state | int(0) > trigger.from_state.state | int(0) }}"
action:
- service: notify.mobile_app_pixel_9_pro
data:
title: "Disk SMART warning"
message: "{{ trigger.to_state.attributes.friendly_name }} increased from {{ trigger.from_state.state }} to {{ trigger.to_state.state }}"
- alias: "Disk SMART check failed"
trigger:
- platform: state
entity_id:
${smartPassedEntitiesYaml}
condition:
- condition: template
value_template: "{{ trigger.to_state.state | lower == 'false' }}"
action:
- service: notify.mobile_app_pixel_9_pro
data:
title: "Disk SMART FAILURE"
message: "{{ trigger.to_state.attributes.friendly_name }} reports SMART failure drive is likely failing"
'';
systemd.services.disk-smart-collect = {
description = "Collect disk SMART data";
serviceConfig = {
Type = "oneshot";
ExecStart = collectScript;
};
};
systemd.timers.disk-smart-collect = {
description = "Periodically collect disk SMART data";
wantedBy = [ "timers.target" ];
timerConfig = {
OnBootSec = "1min";
OnUnitActiveSec = "1min";
};
};
services.nginx.virtualHosts."127.0.0.1" = {
listen = [
{ addr = "127.0.0.1"; port = 9633; }
];
locations."= /smart.json" = {
alias = "${outputDir}/smart.json";
extraConfig = ''
default_type application/json;
'';
};
};
};
}
+188 -141
View File
@@ -6,8 +6,42 @@
}:
let
nginxEnabled = config.mod.nginx.enable;
cfg = config.mod.home-assistant;
script = pkgs.writeShellScript "bt-reset" ''
configFile = pkgs.writeText "ha-configuration.yaml" ''
# Loads default set of integrations. Do not remove.
default_config:
http:
use_x_forwarded_for: true
trusted_proxies:
- 127.0.0.1
# Load frontend themes from the themes folder
frontend:
themes: !include_dir_merge_named themes
automation: !include automations.yaml
script: !include scripts.yaml
scene: !include scenes.yaml
recorder:
purge_keep_days: 365
alert:
fridge_door:
name: Fridge is open
done_message: Fride is closed
entity_id: binary_sensor.kyldorr
state: "on"
repeat: 2
skip_first: true
notifiers:
- mobile_app_pixel_9_pro
${cfg.extraConfig}'';
btResetScript = pkgs.writeShellScript "bt-reset" ''
set -euo pipefail
export PATH="${
lib.makeBinPath [
@@ -62,181 +96,194 @@ let
'';
in
{
mod.homepage.services = [
{
name = "Home Assistant";
port = 8123;
description = "Home automation";
}
];
hardware.bluetooth.enable = true;
virtualisation.oci-containers = {
backend = "podman";
containers.homeassistant = {
image = "ghcr.io/home-assistant/home-assistant:stable";
volumes = [
"/home/alex/.config/home-assistant:/config"
# Pass in bluetooth
"/run/dbus:/run/dbus:ro"
];
environment.TZ = "Europe/Stockholm";
extraOptions = [
"--network=host"
# Allows HA to perform low-level network operations (scan/reset adapter)
"--cap-add=NET_ADMIN"
"--cap-add=NET_RAW"
# Pass in Zigbee antenna
"--device=/dev/serial/by-id/usb-Nabu_Casa_ZBT-2_9C139EAAD464-if00:/dev/ttyACM0"
];
options = {
mod.home-assistant = {
extraConfig = lib.mkOption {
type = lib.types.lines;
default = "";
description = "Extra YAML to append to Home Assistant's configuration.yaml";
};
};
};
services = {
blueman.enable = true;
config = {
mod.homepage.services = [
{
name = "Home Assistant";
port = 8123;
description = "Home automation";
}
];
nginx = lib.mkIf nginxEnabled {
recommendedProxySettings = true;
hardware.bluetooth.enable = true;
virtualHosts."ha.ppp.pm" = {
forceSSL = true;
useACMEHost = "ha.ppp.pm";
virtualisation.oci-containers = {
backend = "podman";
extraConfig = ''
proxy_buffering off;
'';
containers.homeassistant = {
image = "ghcr.io/home-assistant/home-assistant:stable";
locations."/" = {
proxyPass = "http://127.0.0.1:8123";
proxyWebsockets = true;
};
volumes = [
"/home/alex/.config/home-assistant:/config"
"${configFile}:/config/configuration.yaml:ro"
# Pass in bluetooth
"/run/dbus:/run/dbus:ro"
];
environment.TZ = "Europe/Stockholm";
extraOptions = [
"--network=host"
# Allows HA to perform low-level network operations (scan/reset adapter)
"--cap-add=NET_ADMIN"
"--cap-add=NET_RAW"
# Pass in Zigbee antenna
"--device=/dev/serial/by-id/usb-Nabu_Casa_ZBT-2_9C139EAAD464-if00:/dev/ttyACM0"
];
};
};
# Trigger reset via udev when hci0 disappears
udev.extraRules = ''
ACTION=="remove", SUBSYSTEM=="bluetooth", KERNEL=="hci0", \
TAG+="systemd", ENV{SYSTEMD_WANTS}+="bt-reset.service"
'';
};
systemd = {
services = {
# Trigger reset on bluetoothd failure
bluetooth = {
unitConfig.OnFailure = [ "bt-reset.service" ];
};
blueman.enable = true;
bt-reset = {
description = "Reset Bluetooth adapter";
after = [ "bluetooth.service" ];
nginx = lib.mkIf nginxEnabled {
recommendedProxySettings = true;
serviceConfig = {
Type = "oneshot";
ExecStart = script;
virtualHosts."ha.ppp.pm" = {
forceSSL = true;
useACMEHost = "ha.ppp.pm";
Restart = "on-failure";
RestartSec = "10s";
StartLimitIntervalSec = "120";
StartLimitBurst = 3;
};
};
};
extraConfig = ''
proxy_buffering off;
'';
timers.bt-reset = {
description = "Periodically reset Bluetooth adapter";
wantedBy = [ "timers.target" ];
timerConfig = {
OnBootSec = "5min"; # first run 5 min after boot
OnUnitActiveSec = "4h"; # then every 4 hours
RandomizedDelaySec = "5min";
};
};
user = {
timers = {
"update-hetzner-dns" = {
unitConfig = {
Description = "updates Hetzner DNS records";
locations."/" = {
proxyPass = "http://127.0.0.1:8123";
proxyWebsockets = true;
};
timerConfig = {
Unit = "update-hetzner-dns.service";
OnCalendar = "*-*-* *:00/30:00";
Persistent = true;
};
wantedBy = [ "timers.target" ];
};
};
# Trigger reset via udev when hci0 disappears
udev.extraRules = ''
ACTION=="remove", SUBSYSTEM=="bluetooth", KERNEL=="hci0", \
TAG+="systemd", ENV{SYSTEMD_WANTS}+="bt-reset.service"
'';
};
systemd = {
services = {
"update-hetzner-dns" = {
unitConfig = {
Description = "updates Hetzner DNS records";
};
# Trigger reset on bluetoothd failure
bluetooth = {
unitConfig.OnFailure = [ "bt-reset.service" ];
};
bt-reset = {
description = "Reset Bluetooth adapter";
after = [ "bluetooth.service" ];
serviceConfig = {
Type = "exec";
EnvironmentFile = config.age.secrets.hetzner-dns.path;
Type = "oneshot";
ExecStart = btResetScript;
Restart = "on-failure";
RestartSec = "10s";
StartLimitIntervalSec = "120";
StartLimitBurst = 3;
};
};
};
path = [
pkgs.curl
pkgs.coreutils
pkgs.jq
];
timers.bt-reset = {
description = "Periodically reset Bluetooth adapter";
wantedBy = [ "timers.target" ];
timerConfig = {
OnBootSec = "5min"; # first run 5 min after boot
OnUnitActiveSec = "4h"; # then every 4 hours
RandomizedDelaySec = "5min";
};
};
script = ''
SUBDOMAINS="ha komga romm"
INTERFACE="enp3s0"
user = {
timers = {
"update-hetzner-dns" = {
unitConfig = {
Description = "updates Hetzner DNS records";
};
CURRENT_IP=$(curl -s --fail --interface "$INTERFACE" ifconfig.me)
timerConfig = {
Unit = "update-hetzner-dns.service";
OnCalendar = "*-*-* *:00/30:00";
Persistent = true;
};
for SUBDOMAIN in $SUBDOMAINS; do
LAST_IP_FILE="/tmp/hetzner-dns-''${SUBDOMAIN}-ip"
wantedBy = [ "timers.target" ];
};
};
LAST_IP=""
if [[ -f "$LAST_IP_FILE" ]]; then
LAST_IP=$(cat "$LAST_IP_FILE")
fi
services = {
"update-hetzner-dns" = {
unitConfig = {
Description = "updates Hetzner DNS records";
};
if [[ "$CURRENT_IP" == "$LAST_IP" ]]; then
echo "$SUBDOMAIN: IP unchanged, NOOP update."
else
echo "$SUBDOMAIN: Updating IP"
serviceConfig = {
Type = "exec";
EnvironmentFile = config.age.secrets.hetzner-dns.path;
};
JSON_BODY=$(jq -n --arg ip "$CURRENT_IP" '{records: [{value: $ip}]}')
path = [
pkgs.curl
pkgs.coreutils
pkgs.jq
];
curl \
--fail \
-X POST \
-H "Authorization: Bearer $HETZNER_API_TOKEN" \
-H "Content-Type: application/json" \
-d "$JSON_BODY" \
"https://api.hetzner.cloud/v1/zones/ppp.pm/rrsets/''${SUBDOMAIN}/A/actions/set_records" \
&& echo $CURRENT_IP > $LAST_IP_FILE
fi
done
'';
script = ''
SUBDOMAINS="ha komga romm"
INTERFACE="enp3s0"
CURRENT_IP=$(curl -s --fail --interface "$INTERFACE" ifconfig.me)
for SUBDOMAIN in $SUBDOMAINS; do
LAST_IP_FILE="/tmp/hetzner-dns-''${SUBDOMAIN}-ip"
LAST_IP=""
if [[ -f "$LAST_IP_FILE" ]]; then
LAST_IP=$(cat "$LAST_IP_FILE")
fi
if [[ "$CURRENT_IP" == "$LAST_IP" ]]; then
echo "$SUBDOMAIN: IP unchanged, NOOP update."
else
echo "$SUBDOMAIN: Updating IP"
JSON_BODY=$(jq -n --arg ip "$CURRENT_IP" '{records: [{value: $ip}]}')
curl \
--fail \
-X POST \
-H "Authorization: Bearer $HETZNER_API_TOKEN" \
-H "Content-Type: application/json" \
-d "$JSON_BODY" \
"https://api.hetzner.cloud/v1/zones/ppp.pm/rrsets/''${SUBDOMAIN}/A/actions/set_records" \
&& echo $CURRENT_IP > $LAST_IP_FILE
fi
done
'';
};
};
};
};
};
age = {
secrets = {
"hetzner-dns" = {
file = ../../../../secrets/manatee/hetzner-dns.age;
owner = "alex";
group = "users";
age = {
secrets = {
"hetzner-dns" = {
file = ../../../../secrets/manatee/hetzner-dns.age;
owner = "alex";
group = "users";
};
};
};
};