12 Commits

Author SHA1 Message Date
4a79f92c9b Monitor fox via ICMP
It should reply once the machine is connected to the UPC network.
2025-05-29 13:11:12 +02:00
9fc7a3a5d9 Update configuration for UPC network
For now we keep the two IP addresses for BSC and UPC networks, so we
don't lose connectivity.
2025-05-29 13:10:53 +02:00
b0783b8117 Disable home via NFS in fox
It won't be accesible anymore as we won't be in the same LAN.
2025-05-29 13:10:53 +02:00
40cb289b50 Rekey all secrets
Fox is no longer able to use munge or ceph, so we remove the key and
rekey them.
2025-05-29 13:10:53 +02:00
1cbf0cd0b1 Rotate fox SSH host key
Prevent decrypting old secrets by reading the git history.
2025-05-29 13:10:53 +02:00
73ae589303 Distrust fox SSH key
We no longer will share secrets with fox until we can regain our trust.
2025-05-29 13:10:53 +02:00
1e3c2c02f3 Remove Ceph module from fox
It will no longer be accesible from the UPC.
2025-05-29 13:10:53 +02:00
ec0daef96c Remove fox from SLURM 2025-05-29 13:10:53 +02:00
5486d3d6f3 Remove pam_slurm_adopt from fox
We no longer will be able to use SLURM from jungle.
2025-05-29 13:10:53 +02:00
7a2f37aaa2 Add UPC temperature sensor monitoring
These sensors are part of their air quality measurements, which just
happen to be very close to our server room.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-05-29 13:01:37 +02:00
aae6585f66 Add meteocat exporter
Allows us to track ambient temperature changes and estimate the
temperature delta between the server room and exterior temperature.
We should be able to predict when we would need to stop the machines due
to excesive temperature as summer approaches.

Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-05-29 13:01:29 +02:00
1c15e77c83 Add custom nix-daemon exporter
Allows us to see which derivations are being built in realtime. It is a
bit of a hack, but it seems to work. We simply look at the environment
of the child processes of nix-daemon (usually bash) and then look for
the $name variable which should hold the current derivation being
built. Needs root to be able to read the environ file of the different
nix-daemon processes as they are owned by the nixbld* users.

See: https://discourse.nixos.org/t/query-ongoing-builds/23486
Reviewed-by: Aleix Boné <abonerib@bsc.es>
2025-05-29 12:57:07 +02:00
12 changed files with 291 additions and 0 deletions

View File

@@ -3,7 +3,10 @@
{
imports = [
../module/slurm-exporter.nix
../module/meteocat-exporter.nix
../module/upc-qaire-exporter.nix
./gpfs-probe.nix
./nix-daemon-exporter.nix
];
age.secrets.grafanaJungleRobotPassword = {
@@ -108,6 +111,9 @@
"127.0.0.1:${toString config.services.prometheus.exporters.smartctl.port}"
"127.0.0.1:9341" # Slurm exporter
"127.0.0.1:9966" # GPFS custom exporter
"127.0.0.1:9999" # Nix-daemon custom exporter
"127.0.0.1:9929" # Meteocat custom exporter
"127.0.0.1:9928" # UPC Qaire custom exporter
"127.0.0.1:${toString config.services.prometheus.exporters.blackbox.port}"
];
}];
@@ -163,6 +169,7 @@
"8.8.8.8"
"ssfhead"
"anella-bsc.cesca.cat"
"fox.ac.upc.edu"
];
}];
relabel_configs = [

26
m/hut/nix-daemon-builds.sh Executable file
View File

@@ -0,0 +1,26 @@
#!/bin/sh
# Locate nix daemon pid
nd=$(pgrep -o nix-daemon)
# Locate children of nix-daemon
pids1=$(tr ' ' '\n' < "/proc/$nd/task/$nd/children")
# For each children, locate 2nd level children
pids2=$(echo "$pids1" | xargs -I @ /bin/sh -c 'cat /proc/@/task/*/children' | tr ' ' '\n')
cat <<EOF
HTTP/1.1 200 OK
Content-Type: text/plain; version=0.0.4; charset=utf-8; escaping=values
# HELP nix_daemon_build Nix daemon derivation build state.
# TYPE nix_daemon_build gauge
EOF
for pid in $pids2; do
name=$(cat /proc/$pid/environ 2>/dev/null | tr '\0' '\n' | rg "^name=(.+)" - --replace '$1' | tr -dc ' [:alnum:]_\-\.')
user=$(ps -o uname= -p "$pid")
if [ -n "$name" -a -n "$user" ]; then
printf 'nix_daemon_build{user="%s",name="%s"} 1\n' "$user" "$name"
fi
done

View File

@@ -0,0 +1,23 @@
{ pkgs, config, lib, ... }:
let
script = pkgs.runCommand "nix-daemon-exporter.sh" { }
''
cp ${./nix-daemon-builds.sh} $out;
chmod +x $out
''
;
in
{
systemd.services.nix-daemon-exporter = {
description = "Daemon to export nix-daemon metrics";
path = [ pkgs.procps pkgs.ripgrep ];
wantedBy = [ "default.target" ];
serviceConfig = {
Type = "simple";
ExecStart = "${pkgs.socat}/bin/socat TCP4-LISTEN:9999,fork EXEC:${script}";
# Needed root to read the environment, potentially unsafe
User = "root";
Group = "root";
};
};
}

View File

@@ -0,0 +1,17 @@
{ config, lib, pkgs, ... }:
with lib;
{
systemd.services."prometheus-meteocat-exporter" = {
wantedBy = [ "multi-user.target" ];
after = [ "network.target" ];
serviceConfig = {
Restart = mkDefault "always";
PrivateTmp = mkDefault true;
WorkingDirectory = mkDefault "/tmp";
DynamicUser = mkDefault true;
ExecStart = "${pkgs.meteocat-exporter}/bin/meteocat-exporter";
};
};
}

View File

@@ -0,0 +1,17 @@
{ config, lib, pkgs, ... }:
with lib;
{
systemd.services."prometheus-upc-qaire-exporter" = {
wantedBy = [ "multi-user.target" ];
after = [ "network.target" ];
serviceConfig = {
Restart = mkDefault "always";
PrivateTmp = mkDefault true;
WorkingDirectory = mkDefault "/tmp";
DynamicUser = mkDefault true;
ExecStart = "${pkgs.upc-qaire-exporter}/bin/upc-qaire-exporter";
};
};
}

View File

@@ -0,0 +1,25 @@
{ python3Packages, lib }:
python3Packages.buildPythonApplication rec {
pname = "meteocat-exporter";
version = "1.0";
src = ./.;
doCheck = false;
build-system = with python3Packages; [
setuptools
];
dependencies = with python3Packages; [
beautifulsoup4
lxml
prometheus-client
];
meta = with lib; {
description = "MeteoCat Prometheus Exporter";
platforms = platforms.linux;
};
}

View File

@@ -0,0 +1,54 @@
#!/usr/bin/env python3
import time
from prometheus_client import start_http_server, Gauge
from bs4 import BeautifulSoup
from urllib import request
# Configuration -------------------------------------------
meteo_station = "X8" # Barcelona - Zona Universitària
listening_port = 9929
update_period = 60 * 5 # Each 5 min
# ---------------------------------------------------------
metric_tmin = Gauge('meteocat_temp_min', 'Min temperature')
metric_tmax = Gauge('meteocat_temp_max', 'Max temperature')
metric_tavg = Gauge('meteocat_temp_avg', 'Average temperature')
metric_srad = Gauge('meteocat_solar_radiation', 'Solar radiation')
def update(st):
url = 'https://www.meteo.cat/observacions/xema/dades?codi=' + st
response = request.urlopen(url)
data = response.read()
soup = BeautifulSoup(data, 'lxml')
table = soup.find("table", {"class" : "tblperiode"})
rows = table.find_all('tr')
row = rows[-1] # Take the last row
row_data = []
header = row.find('th')
header_text = header.text.strip()
row_data.append(header_text)
for col in row.find_all('td'):
row_data.append(col.text)
try:
# Sometimes it will return '(s/d)' and fail to parse
metric_tavg.set(float(row_data[1]))
metric_tmax.set(float(row_data[2]))
metric_tmin.set(float(row_data[3]))
metric_srad.set(float(row_data[10]))
#print("ok: temp_avg={}".format(float(row_data[1])))
except:
print("cannot parse row: {}".format(row))
metric_tavg.set(float("nan"))
metric_tmax.set(float("nan"))
metric_tmin.set(float("nan"))
metric_srad.set(float("nan"))
if __name__ == '__main__':
start_http_server(port=listening_port, addr="localhost")
while True:
try:
update(meteo_station)
except:
print("update failed")
time.sleep(update_period)

View File

@@ -0,0 +1,11 @@
#!/usr/bin/env python
from setuptools import setup, find_packages
setup(name='meteocat-exporter',
version='1.0',
# Modules to import from other scripts:
packages=find_packages(),
# Executables
scripts=["meteocat-exporter"],
)

View File

@@ -54,4 +54,6 @@ final: prev:
});
prometheus-slurm-exporter = prev.callPackage ./slurm-exporter.nix { };
meteocat-exporter = prev.callPackage ./meteocat-exporter/default.nix { };
upc-qaire-exporter = prev.callPackage ./upc-qaire-exporter/default.nix { };
}

View File

@@ -0,0 +1,24 @@
{ python3Packages, lib }:
python3Packages.buildPythonApplication rec {
pname = "upc-qaire-exporter";
version = "1.0";
src = ./.;
doCheck = false;
build-system = with python3Packages; [
setuptools
];
dependencies = with python3Packages; [
prometheus-client
requests
];
meta = with lib; {
description = "UPC Qaire Prometheus Exporter";
platforms = platforms.linux;
};
}

View File

@@ -0,0 +1,11 @@
#!/usr/bin/env python
from setuptools import setup, find_packages
setup(name='upc-qaire-exporter',
version='1.0',
# Modules to import from other scripts:
packages=find_packages(),
# Executables
scripts=["upc-qaire-exporter"],
)

View File

@@ -0,0 +1,74 @@
#!/usr/bin/env python3
import time
from prometheus_client import start_http_server, Gauge
import requests, json
from datetime import datetime, timedelta
# Configuration -------------------------------------------
listening_port = 9928
update_period = 60 * 5 # Each 5 min
# ---------------------------------------------------------
metric_temp = Gauge('upc_c6_s302_temp', 'UPC C6 S302 temperature sensor')
def genparams():
d = {}
d['topic'] = 'TEMPERATURE'
d['shift_dates_to'] = ''
d['datapoints'] = 301
d['devicesAndColors'] = '1148418@@@#40ACB6'
now = datetime.now()
d['fromDate'] = now.strftime('%d/%m/%Y')
d['toDate'] = now.strftime('%d/%m/%Y')
d['serviceFrequency'] = 'NONE'
# WTF!
for i in range(7):
for j in range(48):
key = 'week.days[{}].hours[{}].value'.format(i, j)
d[key] = 'OPEN'
return d
def measure():
# First we need to load session
s = requests.Session()
r = s.get("https://upc.edu/sirena")
if r.status_code != 200:
print("bad HTTP status code on new session: {}".format(r.status_code))
return
if s.cookies.get("JSESSIONID") is None:
print("cannot get JSESSIONID")
return
# Now we can pull the data
url = "https://upcsirena.app.dexma.com/l_12535/analysis/by_datapoints/data.json"
r = s.post(url, data=genparams())
if r.status_code != 200:
print("bad HTTP status code on data: {}".format(r.status_code))
return
#print(r.text)
j = json.loads(r.content)
# Just take the last one
last = j['data']['chartElementList'][-1]
temp = last['values']['1148418-Temperatura']
return temp
if __name__ == '__main__':
start_http_server(port=listening_port, addr="localhost")
while True:
try:
metric_temp.set(measure())
except:
print("measure failed")
metric_temp.set(float("nan"))
time.sleep(update_period)