Configuration

Once you have successfully installed the Flask-MonitoringDashboard using the instructions from this page, you can use the advanced features by correctly configuring the Dashboard.

Using a configuration file

You can use a single configuration file for all options below. This is explained in the following section. In order to configure the Dashboard with a configuration-file, you can use the following function:

dashboard.config.init_from(file='/<path to file>/config.cfg')

Thus, it becomes:

from flask import Flask
import flask_monitoringdashboard as dashboard

app = Flask(__name__)
dashboard.config.init_from(file='/<path to file>/config.cfg')
# Make sure that you first configure the dashboard, before binding it to your Flask application
dashboard.bind(app)
...

@app.route('/')
def index():
    return 'Hello World!'

if __name__ == '__main__':
  app.run(debug=True)

Instead of having a hard-coded string containing the location of the config file in the code above, it is also possible to define an environment variable that specifies the location of this config file. The line should then be:

dashboard.config.init_from(envvar='DASHBOARD_CONFIG')

This will configure the Dashboard based on the file provided in the environment variable called DASHBOARD_CONFIG.

The content of the configuration file

Once the setup is complete, a configuration file (e.g. ‘config.cfg’) should be set next to the python file that contains the entry point of the app. The following properties can be configured:

[dashboard]
APP_VERSION=1.0
GIT=/<path to your project>/.git/
CUSTOM_LINK='dashboard'
MONITOR_LEVEL=3
OUTLIER_DETECTION_CONSTANT=2.5
SAMPLING_PERIOD=20

[authentication]
USERNAME='admin'
PASSWORD='admin'
GUEST_USERNAME='guest'
GUEST_PASSWORD='[dashboardguest!, second_pw!]'
SECURITY_TOKEN='cc83733cb0af8b884ff6577086b87909'

[database]
TABLE_PREFIX=''
DATABASE=sqlite:////<path to your project>/dashboard.db

[visualization]
TIMEZONE='Europe/Amsterdam'
COLORS={'main':'[0,97,255]',
        'static':'[255,153,0]'}

As can be seen above, the configuration is split into 4 headers:

Dashboard

  • APP_VERSION: The version of the application that you use. Updating the version allows seeing the changes in the execution time of requests over multiple versions.
  • GIT: Since updating the version in the configuration-file when updating code isn’t very convenient, another way is to provide the location of the git-folder. From the git-folder, the version is automatically retrieved by reading the commit-id (hashed value). The specified value is the location to the git-folder. This is relative to the configuration-file.
  • CUSTOM_LINK: The Dashboard can be visited at localhost:5000/{{CUSTOM_LINK}}.
  • MONITOR_LEVEL: The level for monitoring your endpoints. The default value is 3. For more information, see the Rules page.
  • OUTLIER_DETECTION_CONSTANT: When the execution time is greater than \(constant * average\), extra information is logged into the database. A default value for this variable is \(2.5\).
  • SAMPLING_PERIOD: Time between two profiler-samples. The time must be specified in ms. If this value is not set, the profiler monitors continuously.

Authentication

  • USERNAME and PASSWORD: Must be used for logging into the Dashboard. Thus both are required.
  • GUEST_USERNAME and GUEST_PASSWORD: A guest can only see the results, but cannot configure/download any data.
  • SECURITY_TOKEN: The token that is used for exporting the data to other services. If you leave this unchanged, any service is able to retrieve the data from the database.

Database

  • TABLE_PREFIX: A prefix to every table that the Flask-MonitoringDashboard uses, to ensure that there are no conflicts with the other tables, that are specified by the user of the dashboard.
  • DATABASE: Suppose you have multiple projects that you’re working on and want to separate the results. Then you can specify different database_names, such that the result of each project is stored in its own database.

Visualization

  • TIMEZONE: The timezone for converting a UTC timestamp to a local timestamp. For a list of all timezones, use the following:

    import pytz  # pip install pytz
    print(pytz.all_timezones)
    

    The dashboard saves the time of every request by default in a UTC-timestamp. However, if you want to display it in a local timestamp, you need this property.

  • COLORS: The endpoints are automatically hashed into a color. However, if you want to specify a different color for an endpoint, you can set this variable. It must be a dictionary with the endpoint-name as a key, and a list of length 3 with the RGB-values. For example:

    COLORS={'main':'[0,97,255]',
            'static':'[255,153,0]'}
    

What have you configured?

A lot of configuration options, but you might wonder what functionality is now supported in your Flask application? Have a look at this file to find the answer.