Compare commits

...

15 Commits

Author SHA1 Message Date
8d6590364f fix: set model number as additional primary key
Some checks failed
ci / docker (push) Has been cancelled
2025-11-24 17:14:26 +01:00
dd7a81fd0f feat: allow configure database url via env 2025-11-24 17:14:22 +01:00
7b5f28651b update: add build schedule
Some checks failed
ci / docker (push) Failing after 5s
2024-09-27 17:09:55 +02:00
a7f4788291 change: increase temp warning to 60degC 2024-02-05 20:40:55 +01:00
74b48a2477 fix: run as Administratoren to prevent window 2024-01-12 05:35:01 +01:00
18f8436078 fix: broke CSS background on some displays 2024-01-03 17:20:38 +01:00
3df3ddb08e fix: fallback on smart_status for critical bit 2024-01-03 17:05:05 +01:00
72e0210d26 feat: add Perc_Avail_Resrvd_Space to support metrics 2024-01-03 16:55:36 +01:00
edc454f154 feat: windows hourly task snippet 2024-01-03 16:40:10 +01:00
824c108678 fix: handle second_last in first request 2024-01-03 15:16:59 +01:00
08fc17efe0 fix: smart record examples before first request 2024-01-03 14:53:25 +01:00
683ebefbb0 feat: add SMART monitoring support 2024-01-03 14:41:11 +01:00
0842818cbc fix: dont change token on modification 2024-01-03 14:40:57 +01:00
d6ea667733 fix: skip icinga host creation if not configured 2024-01-03 14:39:17 +01:00
935bfa3eef fix: skip icinga connection if not configured 2024-01-03 14:37:50 +01:00
7 changed files with 300 additions and 19 deletions

View File

@@ -4,6 +4,8 @@ on:
push:
branches:
- "master"
schedule:
- cron: "0 2 * * 0"
jobs:
docker:

View File

@@ -41,6 +41,9 @@ def _build_service_name(user, async_service_name):
def create_service(user, async_service_name, app):
if not app.config.get("ICINGA_API_URL"):
return
client = _create_client(app)
name = _build_service_name(user, async_service_name)
host_name = app.config["ASYNC_ICINGA_DUMMY_HOST"]
@@ -67,6 +70,9 @@ def create_service(user, async_service_name, app):
def delete_service(user, async_service_name, app):
if not app.config.get("ICINGA_API_URL"):
return
client = _create_client(app)
name = _build_service_name(user, async_service_name)
host_name = app.config["ASYNC_ICINGA_DUMMY_HOST"]
@@ -83,7 +89,15 @@ def build_icinga_link_for_service(user, service_name, static_configured, app):
url_fmt = "{base}/icingaweb2/monitoring/list/services?service={service}&modifyFilter=1"
name = service_name
return url_fmt.format(base=app.config["ICINGA_WEB_URL"],
host=app.config["ASYNC_ICINGA_DUMMY_HOST"],
icinga_web_url = app.config.get("ICINGA_WEB_URL")
if not icinga_web_url:
icinga_web_url = "ICINGA_WEB_URL_NOT_SET:"
dummy_host=app.config.get("ASYNC_ICINGA_DUMMY_HOST")
if not dummy_host:
dummy_host = "ASYNC_ICINGA_DUMMY_HOST_NOT_SET:"
return url_fmt.format(base=icinga_web_url,
host=dummy_host,
service=name)

162
server.py
View File

@@ -12,7 +12,7 @@ import secrets
import flask_wtf
from flask_wtf import FlaskForm
from wtforms import StringField, SubmitField, BooleanField, DecimalField, HiddenField
from wtforms import StringField, SubmitField, BooleanField, DecimalField, HiddenField, SelectField
from wtforms.validators import DataRequired, Length
from sqlalchemy import Column, Integer, String, Boolean, or_, and_
@@ -24,11 +24,12 @@ from flask_sqlalchemy import SQLAlchemy
from sqlalchemy.sql.expression import func
import icingatools
import smarttools
app = flask.Flask("Icinga Report In Gateway")
app.config['SQLALCHEMY_DATABASE_URI'] = 'sqlite:///database.sqlite'
app.config['SQLALCHEMY_DATABASE_URI'] = os.environ.get('SQLALCHEMY_DATABASE_URI') or 'sqlite:///database.sqlite'
app.config['JSON_CONFIG_FILE'] = 'services.json'
app.config['JSON_CONFIG_DIR'] = 'config'
db = SQLAlchemy(app)
@@ -41,6 +42,7 @@ class Service(db.Model):
token = Column(String)
timeout = Column(Integer)
owner = Column(String)
special_type = Column(String)
staticly_configured = Column(Boolean)
@@ -57,6 +59,22 @@ class Status(db.Model):
dt = datetime.datetime.fromtimestamp(self.timestamp)
return dt.strftime("%d. %B %Y at %H:%M")
class SMARTStatus(db.Model):
__tablename__ = "smart"
service = Column(String, primary_key=True)
timestamp = Column(Integer, primary_key=True)
model_number = Column(String, primary_key=True)
power_cycles = Column(Integer)
temperature = Column(Integer)
available_spare = Column(Integer)
unsafe_shutdowns = Column(Integer)
critical_warning = Column(Integer)
power_cycles = Column(Integer)
power_on_hours = Column(Integer)
wearleveling_count = Column(Integer)
def buildReponseDict(status, service=None):
if not status:
@@ -108,6 +126,7 @@ class EntryForm(FlaskForm):
service = StringField("Service Name")
service_hidden = HiddenField("service_hidden")
special_type = SelectField("Type", choices=["Default", "SMART"])
timeout = DecimalField("Timeout in days", default=30)
def create_entry(form, user):
@@ -116,9 +135,21 @@ def create_entry(form, user):
service_name = form.service.data or form.service_hidden.data
# keep token if modification #
s_tmp = db.session.query(Service).filter(Service.service == service_name).first()
if s_tmp:
token = s_tmp.token
if not token:
raise AssertionError("WTF Service without Token {}".format(service_name))
day_delta = datetime.timedelta(days=int(form.timeout.data))
special_type = form.special_type.data
if form.special_type == "Default":
special_type = None
service = Service(service=service_name, timeout=day_delta.total_seconds(),
owner=user, token=token)
owner=user, token=token, special_type=special_type)
# service.data set = create, service_hidden.data = modify #
if form.service.data:
@@ -148,8 +179,11 @@ def service_details():
icinga_link = icingatools.build_icinga_link_for_service(user, service.service,
service.staticly_configured, app)
smart_entry_list = db.session.query(SMARTStatus).filter(SMARTStatus.service==service.service)
smart_entry = smart_entry_list.order_by(SMARTStatus.timestamp.desc()).first()
return flask.render_template("service_info.html", service=service, flask=flask,
user=user, status_list=status_list, icinga_link=icinga_link)
user=user, status_list=status_list, icinga_link=icinga_link, smart=smart_entry)
@app.route("/entry-form", methods=["GET", "POST", "DELETE"])
@@ -158,7 +192,7 @@ def create_interface():
user = str(flask.request.headers.get("X-Forwarded-Preferred-Username"))
# check if is delete #
operation = flask.request.args.get("operation")
operation = flask.request.args.get("operation")
if operation and operation == "delete" :
service_delete_name = flask.request.args.get("service")
@@ -175,13 +209,14 @@ def create_interface():
return flask.redirect("/overview")
form = EntryForm()
# handle modification #
modify_service_name = flask.request.args.get("service")
if modify_service_name:
service = db.session.query(Service).filter(Service.service == modify_service_name).first()
if service and service.owner == user:
form.service.default = service.service
form.special_type.default = service.special_type
form.timeout.default = datetime.timedelta(seconds=service.timeout).days
form.service_hidden.default = service.service
form.process()
@@ -252,7 +287,7 @@ def default():
if not lastSuccess.timestamp == 0 and delta > timeout and latestInfoIsSuccess:
# lastes info is success but timed out #
lastSuccess.info_text = "Service {} overdue since {}".format(service, str(delta))
lastSuccess.info_text = "Service {} overdue since {}".format(service, str(delta))
if timeout/delta > 0.9 or (delta - timeout) < datetime.timedelta(hours=12):
lastSuccess.status = "WARNING"
else:
@@ -270,16 +305,27 @@ def default():
elif flask.request.method == "POST":
# get variables #
service = flask.request.json["service"]
token = flask.request.json["token"]
status = flask.request.json["status"]
text = flask.request.json["info"]
service = flask.request.json.get("service")
token = flask.request.json.get("token")
status = flask.request.json.get("status")
text = flask.request.json.get("info") or "no_info"
timestamp = datetime.datetime.now().timestamp()
smart = flask.request.json.get("smart")
# check smart json quoting problems #
if smart and type(smart) == str:
try:
smart = json.loads(smart)
except json.decoder.JSONDecodeError as e:
return ("Error in SMART-json {}".format(e), 415)
if not service:
return ("'service' ist empty field in json", 400)
elif not token:
return ("'token' ist empty field in json", 400)
elif not status and not smart:
return ("'status' is empty field in json", 400)
# verify token & service in config #
verifiedServiceObj = db.session.query(Service).filter(
@@ -288,15 +334,94 @@ def default():
if not verifiedServiceObj:
return ("Service ({}) with this token ({}) not found in DB".format(service, token), 401)
else:
status = Status(service=service, timestamp=timestamp, status=status, info_text=text)
# handle a SMART-record submission (with errorhandling) #
if smart and not verifiedServiceObj.special_type == "SMART":
return ("SMART Field for non-SMART type service", 415)
elif smart:
text, status = record_and_check_smart(verifiedServiceObj,
timestamp, smart)
status = Status(service=service, timestamp=timestamp, status=status,
info_text=text)
db.session.merge(status)
db.session.commit()
return ("", 204)
else:
return ("Method not implemented: {}".format(flask.request.method), 405)
def record_and_check_smart(service, timestamp, smart):
if "nvme_smart_health_information_log" in smart:
health_info = smart["nvme_smart_health_information_log"]
else:
health_info = smarttools.normalize(smart)
if not service.special_type == "SMART":
raise AssertionError("Trying to record SMART-record for non-SMART service")
# record the status #
smart_status = SMARTStatus(service=service.service, timestamp=timestamp,
temperature=health_info["temperature"],
critical_warning=health_info["critical_warning"],
unsafe_shutdowns=health_info["unsafe_shutdowns"],
power_cycles=health_info["power_cycles"],
power_on_hours=health_info["power_on_hours"],
available_spare=health_info.get("available_spare"),
model_number=smart.get("model_name"),
wearleveling_count=health_info.get("wearleveling_count"))
db.session.add(smart_status)
db.session.commit()
# check the status #
smart_last_query = db.session.query(SMARTStatus)
smart_last_query = smart_last_query.filter(SMARTStatus.service==service.service)
smart_last = smart_last_query.order_by(sqlalchemy.desc(SMARTStatus.timestamp)).first()
smart_second_last = smart_last_query.order_by(sqlalchemy.desc(
SMARTStatus.timestamp)).offset(1).first()
# last record (max 6 months ago) #
timestampt_minus_6m = datetime.datetime.now() - datetime.timedelta(days=180)
smart_old_query = smart_last_query.filter(
SMARTStatus.timestamp > timestampt_minus_6m.timestamp())
smart_old = smart_old_query.order_by(sqlalchemy.asc(SMARTStatus.timestamp)).first()
# critial != 0 #
if smart_last.critical_warning != 0:
return ("SMART reports disk critical => oO better do something about this", "CRITICAL")
# wearleveling < 20% (SAMSUNG only) #
if smart_last.wearleveling_count and smart_last.wearleveling_count <= 20:
return ("SMART report prefail disk (wear_level < 20%)", "CRITICAL")
# temp max > X #
if smart_last.temperature > 60:
return ("Disk Temperatur {}".format(smart_last.temperature), "CRITICAL")
# available_SSD spare #
spare_change = smart_old.available_spare - smart_last.available_spare
if smart_last.available_spare <= 25:
return ("SSD spare <25 ({}) YOUR DISK WILL DIE SOON".format(spare_change),
"CRITICAL")
elif smart_last.available_spare <= 50:
return ("SSD spare <50 ({})".format(spare_change), "WARNING")
elif spare_change >= 10:
return ("Strong degration in SSD spare space ({} in under 6 months)".format(
spare_change), "WARNING")
# unsafe_shutdowns +1 #
if(smart_second_last and
smart_second_last.unsafe_shutdowns - smart_last.unsafe_shutdowns >= 1):
return ("Disk had {} unsafe shutdowns".format(smart_last.unsafe_shutdowns),
"WARNING")
return ("{} - no problems detected".format(smart_last.model_number), "OK")
def create_app():
db.create_all()
config = {}
@@ -329,14 +454,17 @@ def create_app():
for key in config:
timeout = timeparse.timeparse(config[key]["timeout"])
staticly_configured = True
db.session.merge(Service(service=key, token=config[key]["token"],
db.session.merge(Service(service=key, token=config[key]["token"],
staticly_configured=staticly_configured, timeout=timeout,
owner=config[key]["owner"]))
db.session.commit()
# create dummy host #
icingatools.create_master_host(app)
# create icinga host #
if not app.config.get("ICINGA_API_URL"):
print("ICINGA_API_URL not defined. Not connecting Icinga", file=sys.stderr)
else:
icingatools.create_master_host(app)
if __name__ == "__main__":

54
smarttools.py Normal file
View File

@@ -0,0 +1,54 @@
def normalize(smart):
'''Load different types of SMART outputs'''
ret = dict()
ret.update({ "temperature" : 0 })
ret.update({ "critical_warning" : 0 })
ret.update({ "unsafe_shutdowns" : 0 })
ret.update({ "power_cycles" : 0 })
ret.update({ "power_on_hours" : 0 })
ret.update({ "available_spare" : 100 })
ret.update({ "wearleveling_count" : 100 })
if "ata_smart_attributes" in smart:
# get main table #
table = smart["ata_smart_attributes"]["table"]
# temperatur #
ret["temperature"] = smart["temperature"]["current"]
for el in table:
# look for relevant metrics #
name = el["name"].lower()
target_name = el["name"].lower() # name in return map
# handle value mapping #
use_raw = False
if name == "used_rsvd_blk_cnt_tot":
target_name = "available_spare"
elif name == "power_cylce_count":
target_name = "power_cycles"
use_raw = True
elif name == "power_on_hours":
target_name = "power_on_hours"
use_raw = True
elif name == "perc_avail_resrvd_space":
target_name = "available_spare"
# check if metric should be recorded #
if target_name in ret:
# set return dict #
if use_raw:
value = el["raw"]["value"]
else:
value = el["value"]
ret[target_name] = value
if ret["critical_warning"] == 0 and "smart_status" in smart:
ret["critical_warning"] = int(not smart["smart_status"]["passed"])
return ret

View File

@@ -1,5 +1,6 @@
body{
background: radial-gradient(ellipse at center, #47918a 0%, #0b3161 100%);
background-attachment: fixed;
color: aliceblue !important;
}
@@ -155,6 +156,22 @@ body{
cursor: pointer;
}
.smart-info{
font-family: monospace;
padding-top: 2px;
padding-top: 2px;
padding-left: 5px;
padding-left: 5px;
color: black;
border: none;
outline: none;
cursor: auto;
}
.box{
border-style: solid;
border-width: 1px;

View File

@@ -37,6 +37,9 @@
</br>
{{ form.timeout.label }} {{ form.timeout() }} </br>
</br>
{{ form.special_type.label }} {{ form.special_type() }} </br>
{% if is_modification %}
<input class="form-button mt-4" type="submit" value="Send Modification">
{% else %}

View File

@@ -71,8 +71,39 @@
class="service-token">Secret Token: {{ service.token }}</div>
</div>
{% if service.special_type == "SMART" %}
<div class="clear smart-info mt-3" style="background-color: orange;">
Smart Monitor {% if smart %} for: {{ smart.model_number }} {% endif %}
</div>
<div class="clear smart-info mt-3" style="background-color: orange;">
Example below requires smartmontools ("smartctl") in PATH.
On Linux this is usually available via the package manager,
on Windows install it from the
<a style="text-decoration: underline; color: #5000e1; font-weight: bold;" href="https://www.smartmontools.org/wiki/Download#InstalltheWindowspackage">offical page</a>.
</div>
{% endif %}
{% if service.special_type == "SMART" %}
<h5 class="clear my-4">Linux</h5>
{% else %}
<h5 class="clear my-4">Curl</h5>
{% endif %}
<div class="ml-3 example">
{% if service.special_type == "SMART" %}
SMART='{ <br>
<div class="example-indent">
"service" : "{{ service.service }}", <br>
"token" : "{{ service.token }}", <br>
"status" : "N/A", <br>
"smart" : '$(/sbin/smartctl -a /dev/nvme0n1 --json)' <br>
</div>
}' <br><br>
curl -X POST -H "Content-Type: application/json" \ <br>
<div class="example-indent">
--data "${SMART}" \ <br>
{{ flask.request.url_root.replace("http://", "https://" )}}report
</div>
{% else %}
curl -X POST \ <br>
<div class="example-indent">
-H "Content-Type: application/json" \ <br>
@@ -81,8 +112,39 @@
"status" : "OK", "info" : "Free Text Information here" }' \<br>
{{ flask.request.url_root.replace("http://", "https://" )}}report
</div>
{% endif %}
</div>
{% if service.special_type == "SMART" %}
<h5 class="my-4">Windows</h5>
<div class="ml-3 example">
$SMART = @{ <br>
<div class="example-indent">
service = "{{ service.service }}"<br>
token = "{{ service.token }}"<br>
status = "N/A"<br>
smart = "$(smartctl -a C: --json | Out-String)"<br>
</div>
} | ConvertTo-Json<br><br>
Invoke-RestMethod -TimeoutSec 2 -Uri "{{ flask.request.url_root.replace("http://", "https://" )}}report" -Method Post -Headers @{"Content-Type"="application/json"} -Body $SMART
</div>
<!-- register task example -->
<h5 class="my-4">Windows Task (requires Admin-Powershell)</h5>
<div class="ml-3 example">
$ScriptPath = Join-Path $HOME -ChildPath "smart_monitor.ps1" <br>
echo '$SMART = @{ <br>
<div class="example-indent">
service = "{{ service.service }}"<br>
token = "{{ service.token }}"<br>
status = "N/A"<br>
smart = "$(smartctl -a C: --json | Out-String)"<br>
</div>
} | ConvertTo-Json<br><br>
Invoke-RestMethod -TimeoutSec 2 -Uri "{{ flask.request.url_root.replace("http://", "https://" )}}report" -Method Post -Headers @{"Content-Type"="application/json"} -Body $SMART' &gt; $ScriptPath <br>
schtasks /create /tn SMART_Monitor /tr "powershell.exe -executionpolicy bypass -File '$ScriptPath'" /sc hourly /mo 1 /ru "Administratoren"<br>
echo "Done" <br>
</div>
{% else %}
<h5 class="my-4">Python</h5>
<div class="ml-3 example">
import requests<br>
@@ -97,6 +159,7 @@
</div>
</div>
</div>
{% endif %}
<table class="mb-4 mt-5 status-table">
<thead>