OpenWISP 2 monitoring module (Work in progress).
Table of Contents:
- Available Features
- Install Dependencies
- Setup (integrate in an existing Django project)
- Device Health Status
- Settings
OPENWISP_MONITORING_SHORT_RETENTION_POLICY
OPENWISP_MONITORING_AUTO_PING
OPENWISP_MONITORING_AUTO_CHARTS
OPENWISP_MONITORING_CRITICAL_DEVICE_METRICS
OPENWISP_MONITORING_HEALTH_STATUS_LABELS
OPENWISP_MONITORING_MANAGEMENT_IP_ONLY
OPENWISP_MONITORING_DEVICE_RECOVERY_DETECTION
OPENWISP_MONITORING_CHARTS
OPENWISP_MONITORING_MAC_VENDOR_DETECTION
- Signals
- Default Alerts / Notifications
- Installing for development
- Extending openwisp-monitoring
- 1. Initialize your custom module
- 2. Install
openwisp-monitoring
- 3. Add
EXTENDED_APPS
- 4. Add
openwisp_utils.staticfiles.DependencyFinder
- 5. Add
openwisp_utils.loaders.DependencyLoader
- 6. Inherit the AppConfig class
- 7. Create your custom models
- 8. Add swapper configurations
- 9. Create database migrations
- 10. Create your custom admin
- 11. Create root URL configuration
- 12. Create celery.py
- 13. Import Celery Tasks
- 14. Create the custom command
run_checks
- 15. Import the automated tests
- Other base classes that can be inherited and extended
- Registering new notification types
- Contributing
- Collects and displays device status information like uptime, RAM status, CPU load averages, Interface addresses, WiFi interface status and associated clients, Neighbors information, DHCP Leases, Disk/Flash status
- Collection of monitoring information in a timeseries database (currently only influxdb is supported)
- Monitoring charts for uptime, packet loss, round trip time (latency), associated wifi clients, interface traffic, RAM usage, CPU load, flash/disk usage
- Charts can be viewed at resolutions of 1 day, 3 days, a week, a month and a year
- CSV Export of monitoring data
- Possibility to define custom Charts
- Extensible active check system: it's possible to write additional checks that are run periodically using python classes
- Configurable alerts and web notifications
- API to retrieve the chart metrics and status information of each device
We use InfluxDB to store metrics and Redis as celery broker (you can use a different broker if you want). The recommended way for development is running them using Docker so you will need to install docker and docker-compose beforehand.
In case you prefer not to use Docker you can install InfluxDB and Redis from your repositories, but keep in mind that the version packaged by your distribution may be different.
Install spatialite and sqlite:
sudo apt-get install sqlite3 libsqlite3-dev openssl libssl-dev
sudo apt-get install gdal-bin libproj-dev libgeos-dev libspatialite-dev
Optionally, install fping
if you need to use the ping active check:
sudo apt install -y fping
Follow the setup instructions of openwisp-controller, then add the settings described below.
INSTALLED_APPS = [
# django apps
# openwisp2 admin theme (must be loaded here)
'openwisp_utils.admin_theme',
# all-auth
'django.contrib.sites',
'allauth',
'allauth.account',
'allauth.socialaccount',
'django_extensions',
# openwisp2 modules
'openwisp_users',
'openwisp_controller.pki',
'openwisp_controller.config',
'openwisp_controller.connection',
# monitoring
'notifications',
'openwisp_monitoring.monitoring',
'openwisp_monitoring.device',
'openwisp_monitoring.check',
# admin
'django.contrib.admin',
'django.forms',
# other dependencies ...
]
# Make sure you change them in production
# You can select one of the backends located in openwisp_monitoring.db.backends
TIMESERIES_DATABASE = {
'BACKEND': 'openwisp_monitoring.db.backends.influxdb',
'USER': 'openwisp',
'PASSWORD': 'openwisp',
'NAME': 'openwisp2',
'HOST': 'localhost',
'PORT': '8086',
}
urls.py
:
from django.conf import settings
from django.conf.urls import include, url
from django.contrib.staticfiles.urls import staticfiles_urlpatterns
from openwisp_utils.admin_theme.admin import admin, openwisp_admin
openwisp_admin()
urlpatterns = [
url(r'^admin/', include(admin.site.urls)),
url(r'', include('openwisp_controller.urls')),
url(r'', include('openwisp_monitoring.urls')),
]
urlpatterns += staticfiles_urlpatterns()
Add apptemplates.Loader to template loaders:
TEMPLATES = [
{
'BACKEND': 'django.template.backends.django.DjangoTemplates',
'DIRS': [os.path.join(os.path.dirname(BASE_DIR), 'templates')],
'OPTIONS': {
'loaders': [
'apptemplates.Loader',
'django.template.loaders.filesystem.Loader',
'django.template.loaders.app_directories.Loader',
'openwisp_utils.loaders.DependencyLoader',
],
'context_processors': [
'django.template.context_processors.debug',
'django.template.context_processors.request',
'django.contrib.auth.context_processors.auth',
'django.contrib.messages.context_processors.messages',
],
},
}
]
Configure caching (you may use a different cache storage if you want):
CACHES = {
'default': {
'BACKEND': 'django_redis.cache.RedisCache',
'LOCATION': 'redis://localhost/0',
'OPTIONS': {
'CLIENT_CLASS': 'django_redis.client.DefaultClient',
}
}
}
SESSION_ENGINE = 'django.contrib.sessions.backends.cache'
SESSION_CACHE_ALIAS = 'default'
Configure celery (you may use a different broker if you want):
# here we show how to configure celery with redis but you can
# use other brokers if you want, consult the celery docs
CELERY_BROKER_URL = 'redis://localhost/1'
CELERY_BEAT_SCHEDULE = {
'run_checks': {
'task': 'openwisp_monitoring.check.tasks.run_check',
'schedule': timedelta(minutes=5),
},
}
INSTALLED_APPS.append('djcelery_email')
EMAIL_BACKEND = 'djcelery_email.backends.CeleryEmailBackend'
If you decide to use redis (as shown in these examples), install the requierd python packages:
pip install redis django-redis
The possible values for the health status field (DeviceMonitoring.status
)
are explained below.
Whenever a new device is created it will have UNKNOWN
as it's default Heath Status.
It implies that the system doesn't know whether the device is reachable yet.
Everything is working normally.
One of the metrics has a value which is not in the expected range (the threshold value set in the alert settings has been crossed).
Example: CPU usage should be less than 90% but current value is at 95%.
One of the metrics defined in OPENWISP_MONITORING_CRITICAL_DEVICE_METRICS
has a value which is not in the expected range
(the threshold value set in the alert settings has been crossed).
Example: ping is by default a critical metric which is expected to be always 1 (reachable).
type: | str |
default: | 24h0m0s |
The default retention policy used to store raw device data.
This data is only used to assess the recent status of devices, keeping it for a long time would not add much benefit and would cost a lot more in terms of disk space.
type: | bool |
default: | True |
Whether ping checks are created automatically for devices.
type: | list |
default: | ('traffic', 'wifi_clients', 'uptime', 'packet_loss', 'rtt') |
Automatically created charts.
type: | list of dict objects |
default: | [{'key': 'ping', 'field_name': 'reachable'}] |
Device metrics that are considered critical:
when a value crosses the boundary defined in the "threshold value" field
of the alert settings related to one of these metric types, the health status
of the device related to the metric moves into CRITICAL
.
By default, if devices are not reachable by pings they are flagged as CRITICAL
.
type: | dict |
default: | {'unknown': 'unknown', 'ok': 'ok', 'problem': 'problem', 'critical': 'critical'} |
This setting allows to change the health status labels, for example, if we
want to use online
instead of ok
and offline
instead of critical
,
you can use the following configuration:
OPENWISP_MONITORING_HEALTH_STATUS_LABELS = {
'ok': 'online',
'problem': 'problem',
'critical': 'offline'
}
type: | bool |
default: | True |
By default, only the management IP will be used to perform active checks to the devices.
If the devices are connecting to your OpenWISP instance using a shared layer2
network, hence the OpenWSP server can reach the devices using the last_ip
field, you can set this to False
.
type: | bool |
default: | True |
When device recovery detection is enabled, recoveries are discovered as soon as a device contacts the openwisp system again (eg: to get the configuration checksum or to send monitoring metrics).
This feature is enabled by default.
type: | dict |
default: | {} |
This setting allows to define additional charts or to override
the default chart configuration defined in
openwisp_monitoring.monitoring.charts.DEFAULT_CHARTS
.
For example, if you want to change the traffic chart to show MB (megabytes) instead of GB (Gigabytes) you can use:
OPENWISP_MONITORING_CHARTS = {
'traffic': {
'unit': ' MB',
'description': (
'Network traffic, download and upload, measured on '
'the interface "{metric.key}", measured in MB.'
),
'query': {
'influxdb': (
"SELECT SUM(tx_bytes) / 1000000 AS upload, "
"SUM(rx_bytes) / 1000000 AS download FROM {key} "
"WHERE time >= '{time}' AND content_type = '{content_type}' "
"AND object_id = '{object_id}' GROUP BY time(1d)"
)
},
}
}
Or if you want to define a new chart configuration, which you can then call in your custom code (eg: a custom check class), you can do so as follows:
from django.utils.translation import gettext_lazy as _
OPENWISP_MONITORING_CHARTS = {
'ram': {
'type': 'line',
'title': 'RAM usage',
'description': 'RAM usage',
'unit': 'bytes',
'order': 100,
'query': {
'influxdb': (
"SELECT MEAN(total) AS total, MEAN(free) AS free, "
"MEAN(buffered) AS buffered FROM {key} WHERE time >= '{time}' AND "
"content_type = '{content_type}' AND object_id = '{object_id}' "
"GROUP BY time(1d)"
)
},
}
}
In case you just want to change the colors used in a chart here's how to do it:
OPENWISP_MONITORING_CHARTS = {
'traffic': {
'colors': ['#000000', '#cccccc']
}
}
type: | bool |
default: | True |
Indicates whether mac addresses will be complemented with hardware vendor information by performing lookups on the OUI (Organization Unique Identifier) table.
This feature is enabled by default.
Path: openwisp_monitoring.device.signals.device_metrics_received
Arguments:
instance
: instance ofDevice
whose metrics have been receivedrequest
: the HTTP request object
This signal is emitted when device metrics are received to the DeviceMetric
view (only when using HTTP POST).
The signal is emitted just before a successful response is returned, it is not sent if the response was not successful.
Path: openwisp_monitoring.monitoring.signals.threshold_crossed
Arguments:
sender
: Metric classmetric
:Metric
object whose threshold defined in related alert settings was crossedalert_settings
:AlertSettings
related to theMetric
target
: relatedDevice
objectfirst_time
: it will be set to true when the metric is written for the first time. It shall be set to false afterwards.
first_time
parameter can be used to avoid initiating unneeded actions.
For example, sending recovery notifications.
This signal is emitted when the threshold value of a Metric
defined in
alert settings is crossed.
Notification Type | Use |
threshold_crossed |
Fires when a metric crosses the boundary defined in the threshold value of the alert settings. |
threshold_recovery |
Fires when a metric goes back within the expected range. |
connection_is_working |
Fires when the connection to a device is working. |
connection_is_not_working |
Fires when the connection (eg: SSH) to a device stops working (eg: credentials are outdated, management IP address is outdated, or device is not reachable). |
Install your forked repo:
git clone git://github.com/<your_fork>/openwisp-monitoring
cd openwisp-monitoring/
python setup.py develop
Install test requirements:
pip install -r requirements-test.txt
Start Redis and InfluxDB using docker-compose:
docker-compose up -d
Create the Django database:
cd tests/
./manage.py migrate
./manage.py createsuperuser
Launch development server:
./manage.py runserver 0.0.0.0:8000
You can access the admin interface at http://127.0.0.1:8000/admin/.
Run celery and celery-beat with the following commands (separate terminal windows are needed):
# (cd tests)
celery -A openwisp2 worker -l info
celery -A openwisp2 beat -l info
Run tests with:
# run qa checks
./run-qa-checks
# standard tests
./runtests.py
# tests for the sample app
SAMPLE_APP=1 ./runtests.py
When running the last line of the previous example, the environment variable
SAMPLE_APP
activates the sample apps in /tests/openwisp2/
which are simple django apps that extend openwisp-monitoring
with
the sole purpose of testing its extensibility, for more information regarding
this concept, read the following section.
One of the core values of the OpenWISP project is Software Reusability, for this reason openwisp-monitoring provides a set of base classes which can be imported, extended and reused to create derivative apps.
In order to implement your custom version of openwisp-monitoring, you need to perform the steps described in the rest of this section.
When in doubt, the code in the test project
and the sample apps
namely sample_check,
sample_monitoring, sample_device_monitoring
will guide you in the correct direction:
just replicate and adapt that code to get a basic derivative of
openwisp-monitoring working.
Premise: if you plan on using a customized version of this module, we suggest to start with it since the beginning, because migrating your data from the default module to your extended version may be time consuming.
The first thing you need to do in order to extend any openwisp-monitoring app is create a new django app which will contain your custom version of that openwisp-monitoring app.
A django app is nothing more than a
python package
(a directory of python scripts), in the following examples we'll call these django apps as
mycheck
, mydevicemonitoring
, mymonitoring
but you can name it how you want:
django-admin startapp mycheck django-admin startapp mydevicemonitoring django-admin startapp mymonitoring
Keep in mind that the command mentioned above must be called from a directory which is available in your PYTHON_PATH so that you can then import the result into your project.
Now you need to add mycheck
to INSTALLED_APPS
in your settings.py
,
ensuring also that openwisp_monitoring.check
has been removed:
INSTALLED_APPS = [
# ... other apps ...
# 'openwisp_monitoring.check', <-- comment out or delete this line
# 'openwisp_monitoring.device', <-- comment out or delete this line
# 'openwisp_monitoring.monitoring' <-- comment out or delete this line
'mycheck',
'mydevicemonitoring',
'mymonitoring',
]
For more information about how to work with django projects and django apps, please refer to the django documentation.
Install (and add to the requirement of your project) openwisp-monitoring:
pip install --U https://github.com/openwisp/openwisp-monitoring/tarball/master
Add the following to your settings.py
:
EXTENDED_APPS = ['device_monitoring', 'monitoring', 'check']
Add openwisp_utils.staticfiles.DependencyFinder
to
STATICFILES_FINDERS
in your settings.py
:
STATICFILES_FINDERS = [
'django.contrib.staticfiles.finders.FileSystemFinder',
'django.contrib.staticfiles.finders.AppDirectoriesFinder',
'openwisp_utils.staticfiles.DependencyFinder',
]
Add openwisp_utils.loaders.DependencyLoader
to TEMPLATES
in your settings.py
:
TEMPLATES = [
{
'BACKEND': 'django.template.backends.django.DjangoTemplates',
'OPTIONS': {
'loaders': [
'django.template.loaders.filesystem.Loader',
'django.template.loaders.app_directories.Loader',
'openwisp_utils.loaders.DependencyLoader',
],
'context_processors': [
'django.template.context_processors.debug',
'django.template.context_processors.request',
'django.contrib.auth.context_processors.auth',
'django.contrib.messages.context_processors.messages',
],
},
}
]
Please refer to the following files in the sample app of the test project:
- sample_check/__init__.py.
- sample_check/apps.py.
- sample_monitoring/__init__.py.
- sample_monitoring/apps.py.
- sample_device_monitoring/__init__.py.
- sample_device_monitoring/apps.py.
For more information regarding the concept of AppConfig
please refer to
the "Applications" section in the django documentation.
To extend check
app, refer to sample_check models.py file.
To extend monitoring
app, refer to sample_monitoring models.py file.
To extend device_monitoring
app, refer to sample_device_monitoring models.py file.
Note:
- For doubts regarding how to use, extend or develop models please refer to the "Models" section in the django documentation.
- For doubts regarding proxy models please refer to proxy models.
Add the following to your settings.py
:
# Setting models for swapper module
# For extending check app
CHECK_CHECK_MODEL = 'YOUR_MODULE_NAME.Check'
# For extending monitoring app
MONITORING_CHART_MODEL = 'YOUR_MODULE_NAME.Chart'
MONITORING_METRIC_MODEL = 'YOUR_MODULE_NAME.Metric'
MONITORING_ALERTSETTINGS_MODEL = 'YOUR_MODULE_NAME.AlertSettings'
# For extending device_monitoring app
DEVICE_MONITORING_DEVICEDATA_MODEL = 'YOUR_MODULE_NAME.DeviceData'
DEVICE_MONITORING_DEVICEMONITORING_MODEL = 'YOUR_MODULE_NAME.DeviceMonitoring'
Substitute <YOUR_MODULE_NAME>
with your actual django app name
(also known as app_label
).
Create and apply database migrations:
./manage.py makemigrations ./manage.py migrate
For more information, refer to the "Migrations" section in the django documentation.
To extend check
app, refer to sample_check admin.py file.
To extend monitoring
app, refer to sample_monitoring admin.py file.
To extend device_monitoring
app, refer to sample_device_monitoring admin.py file.
To introduce changes to the admin, you can do it in the two ways described below.
Note: for doubts regarding how the django admin works, or how it can be customized, please refer to "The django admin site" section in the django documentation.
If the changes you need to add are relatively small, you can resort to monkey patching.
For example, for check
app you can do it as:
from openwisp_monitoring.check.admin import CheckAdmin
CheckAdmin.list_display.insert(1, 'my_custom_field')
CheckAdmin.ordering = ['-my_custom_field']
Similarly for device_monitoring
app, you can do it as:
from openwisp_monitoring.device.admin import DeviceAdmin
DeviceAdmin.list_display.insert(1, 'my_custom_field')
DeviceAdmin.ordering = ['-my_custom_field']
Similarly for monitoring
app, you can do it as:
from openwisp_monitoring.monitoring.admin import MetricAdmin, AlertSettingsAdmin
MetricAdmin.list_display.insert(1, 'my_custom_field')
MetricAdmin.ordering = ['-my_custom_field']
AlertSettingsAdmin.list_display.insert(1, 'my_custom_field')
AlertSettingsAdmin.ordering = ['-my_custom_field']
If you need to introduce significant changes and/or you don't want to resort to monkey patching, you can proceed as follows:
For check
app,
from django.contrib import admin
from openwisp_monitoring.check.admin import CheckAdmin as BaseCheckAdmin
from swapper import load_model
Check = load_model('Check')
admin.site.unregister(Check)
@admin.register(Check)
class CheckAdmin(BaseCheckAdmin):
# add your changes here
For device_monitoring
app,
from django.contrib import admin
from openwisp_monitoring.device_monitoring.admin import DeviceAdmin as BaseDeviceAdmin
from openwisp_controller.config.models import Device
admin.site.unregister(Device)
@admin.register(Device)
class DeviceAdmin(BaseDeviceAdmin):
# add your changes here
For monitoring
app,
from django.contrib import admin
from openwisp_monitoring.monitoring.admin import (
AlertSettingsAdmin as BaseAlertSettingsAdmin,
MetricAdmin as BaseMetricAdmin
)
from swapper import load_model
Metric = load_model('Metric')
AlertSettings = load_model('AlertSettings')
admin.site.unregister(Metric)
admin.site.unregister(AlertSettings)
@admin.register(Metric)
class MetricAdmin(BaseMetricAdmin):
# add your changes here
@admin.register(AlertSettings)
class AlertSettingsAdmin(BaseAlertSettingsAdmin):
# add your changes here
Please refer to the urls.py file in the test project.
For more information about URL configuration in django, please refer to the "URL dispatcher" section in the django documentation.
Please refer to the celery.py file in the test project.
For more information about the usage of celery in django, please refer to the "First steps with Django" section in the celery documentation.
Add the following in your settings.py to import celery tasks from device_monitoring
app.
CELERY_IMPORTS = ('openwisp_monitoring.device.tasks',)
Please refer to the run_checks.py file in the test project.
For more information about the usage of custom management commands in django, please refer to the "Writing custom django-admin commands" section in the django documentation.
When developing a custom application based on this module, it's a good idea to import and run the base tests too, so that you can be sure the changes you're introducing are not breaking some of the existing features of openwisp-monitoring.
In case you need to add breaking changes, you can overwrite the tests defined in the base classes to test your own behavior.
For, extending check
app see the tests of sample_check app
to find out how to do this.
For, extending device_monitoring
app see the tests of sample_device_monitoring app
to find out how to do this.
For, extending monitoring
app see the tests of sample_monitoring app
to find out how to do this.
The following steps are not required and are intended for more advanced customization.
This view is responsible for displaying Charts
and Status
primarily.
The full python path is: openwisp_monitoring.device.api.views.DeviceMetricView
.
If you want to extend this view, you will have to perform the additional steps below.
Step 1. Import and extend view:
# mydevice/api/views.py
from openwisp_monitoring.device.api.views import (
DeviceMetricView as BaseDeviceMetricView
)
class DeviceMetricView(BaseDeviceMetricView):
# add your customizations here ...
pass
Step 2: remove the following line from your root ``urls.py`` file:
url(
r'^api/v1/monitoring/device/(?P<pk>[^/]+)/$',
views.device_metric,
name='api_device_metric',
),
Step 3: add an URL route pointing to your custom view in ``urls.py`` file:
# urls.py
from mydevice.api.views import DeviceMetricView
urlpatterns = [
# ... other URLs
url(r'^(?P<path>.*)$', DeviceMetricView.as_view(), name='api_device_metric',),
]
You can define your own notification types using register_notification_type
function from OpenWISP
Notifications. For more information, see the relevant openwisp-notifications section about registering notification types.
Once a new notification type is registered, you have to use the "notify" signal provided in openwisp-notifications to send notifications for this type.
Please refer to the OpenWISP contributing guidelines.