Skip to content

Latest commit

 

History

History
252 lines (190 loc) · 10.4 KB

INTRODUCTION.md

File metadata and controls

252 lines (190 loc) · 10.4 KB
Raw

Webviz introduction

Fundamental configuration

A configuration consists of some mandatory properties, e.g. app title, and one or more pages. A page has a title and some content. Each page can contain as many plugins as you want.

Plugins represent predefined content, which can take one or more arguments. Lists and descriptions of installed plugins can be found on the other subpages.

Content which is not plugins is interpreted as text paragraphs.

A simple example configuration:

# This is a webviz configuration file example.
# The configuration files use the YAML standard (https://en.wikipedia.org/wiki/YAML).

title: Reek Webviz Demonstration

pages:

 - title: Front page
   content:
    - BannerImage:
        image: ./example_banner.png
        title: My banner image
    - Webviz created from a configuration file.

 - title: Markdown example
   content:
    - Markdown:
        markdown_file: ./example-markdown.md

Command line usage

Get documentation

You can always run webviz --help to see available command line options. To see command line options on a subcommand, run e.g. webviz build --help.

📚 To open the webviz documentation on all installed plugins, run webviz docs.

Portable vs. non-portable

Assuming you have a configuration file your_config.yml, there are two main usages of webviz:

webviz build your_config.yml

and

webviz build your_config.yml --portable ./some_output_folder
python ./some_output_folder/webviz_app.py

Portable

The portable way is useful when one or more plugins included in the configuration need to do some time-consuming data aggregation on their own, before presenting it to the user. The time-consuming part will then be done in the build step, and you can run your created application as many time as you want afterwards, with as little waiting time as possible.

The --portable way also has the benefit of creating a 🐳 Docker setup for your application - ready to be deployed to e.g. a cloud provider.

Non-portable

Non-portable is the easiest way if none of the plugins have time-consuming data aggregation to do.

A feature in Dash, used by webviz is hot reload. When the Dash Python code file is saved, the content seen in the web browser is automatically reloaded (no need for localhost server restart). This feature is passed on to the Webviz configuration utility, meaning that if you run

webviz build ./examples/basic_example.yaml

and then modify ./examples/basic_example.yaml while the Webviz application is still running, a hot reload will occur.

Localhost HSTS

Previous versions of webviz generated a local certificate to force localhost connections to go through HTTPS. This is no longer the case and localhost connections use HTTP. As such, the webviz certificate command has been deprecated.

Some browsers will force HTTPS and require extra steps to remove this security. Note that this is safe as no external computer may connect to a localhost server.

If you're having issues connecting to a localhost server running Webviz due to security issues, perform the following steps:

Google Chrome and Chromium

These are the steps to remove HSTS, a security feature that forces HTTPS connections even though the user has specified HTTP:

  1. Navigate to chrome://net-internals/#hsts
  2. In the Delete domain security policies, type in "localhost" and click delete
Firefox

Firefox does not have issues connecting to localhost addresses over HTTP.

User preferences

You can set preferred 🌈 theme and/or 🌍 browser, such that webviz remembers it for later runs. E.g.

webviz preferences --theme equinor --browser firefox

YAML schema

If you are using an editor that supports YAML file validation towards a schema (like Visual Studio Code), Webviz can provide your editor with a schema which then will help you with e.g. auto-completion, detect mistakes immediately and get hover description on different plugins.

Visual Studio Code

If you are using Visual Studio Code as editor you can follow these steps in order to enable validation:

  1. Open Visual Studio Code by e.g. running the command code
  2. From there, open the "Extension Marketplace" by e.g. Ctrl+Shift+X
  3. Search for the extension redhat.vscode-yaml (which is Red Hat's YAML extension)
  4. Click Install
  5. Open settings.json by e.g. Ctrl+,
  6. Search for the setting Yaml: Schemas and click "Edit in settings.json"
  7. Between yaml.schemas: { and the following } insert: 
    "PATH_TO_YOUR_SCHEMA": ["*webviz*.{yml, yaml}"]
    where you replace PATH_TO_YOUR_SCHEMA with what you get from one of these options:
    • Use one of the online URLs where regularly updated schemas are hosted.
    • Run webviz schema and use the file path the schema is written to. Note that if you use this option you need to rerun webviz schema when you update or install new plugin projects in order to keep the schema used by the editor updated with what is actually installed.

After these steps you will start getting help from the editor on YAML files having webviz as part of the filename. You might have to restart the editor after updating the settings.

Logging

The default logging configuration for a Webviz application is to only show log messages with level WARNING or ERROR. This global logging level can be changed on the command line using the --loglevel argument like this:

webviz build ./examples/basic_example.yaml --loglevel INFO

In practice, setting the global logging level to INFO or DEBUG will likely flood the log with messages from various sub-modules, possibly obscuring messages from the module(s) of interest. To allow finer grain control over logging, Webviz accepts the --logconfig command line argument which allows detailed logging configuration through a YAML file:

webviz build ./examples/basic_example.yaml --logconfig my_log_config.yaml

The YAML file is expected to contain a dictionary with logging configuration adhering to the schema described here. This allows for full flexibility with regards to configuring logging in the Webviz application, including setting multiple handlers, filtering and customized log message formatting.

Included below is a simple YAML dictionary showing an incremental configuration suitable for controlling the level of the root logger and a few named loggers. Please note the mandatory version key and the inclusion of the incremental key (see Python docs for more information on incremental configurations). If you need more flexibility than incremental configurations allow, you are free to specify a full configuration, but be aware that this quickly becomes quite involved and requires a good understanding of the schema mentioned above.

# This is a skeleton for a simple, incremental logging configuration.
# See https://docs.python.org/3/library/logging.config.html#logging-config-dictschema for schema.

version: 1
incremental: True

root:
  level: DEBUG
loggers:
  werkzeug:
    level: ERROR
  some.other.module:
    level: INFO

Please note that if both --loglevel and --logconfig are specified, the latter will take precedence. Internally, the global log level set by --loglevel will be applied first. Then the configuration specified by --logconfig will be applied, possibly overwriting any overlapping settings.

Deployment

When you have created a portable Webviz application, you are approximately one command away of either creating a new application in cloud (or updating an existing application). Note that all deploy workflows listed below require that you install the deployment dependencies first by running

pip install webviz-config[deployment]
Azure

Automatic deployment to Azure Web app service is very much possible, and only a community PR away (most of the machinery and Azure CLI wrappers are already in place through the Radix deployment feature).

Radix

Radix is an open source hosting framework by Equinor, built on top of Kubernetes.

Initial deploy

For the initial deploy workflow, you will need to have the GitHub CLI binary gh and Radix CLI binary rx in your $PATH. You can download them from https://github.com/cli/cli/releases and https://github.com/equinor/radix-cli/releases respectively. Also make sure you have a somewhat recent git version available (>=2.0).

When you execute

webviz deploy radix ./your_portable_app owner/reponame --initial-deploy

where ./your_portable_app is a generated portable app, Webviz will do the following:

  • Guide you through Azure CLI, GitHub CLI and Radix CLI login procedure (if not already logged in).
  • Ask for your input on necessary Radix and Azure configuration settings.
  • Create Azure app registration (with necessary callback setting, client secret etc).
  • Create Azure blob storage account and container (if it doesn't already exist).
  • Upload static app resources (i.e. those coming from webviz_storage) to the blob container.
  • Create a private GitHub repository at https://github.com/owner/reponame.
  • Add the Python code to the repository, together with automatically generated Radix configuration file.
  • Turning on GitHub vulnerability alerts for the new repository.
  • Create the new Radix application, and configure a webhook from the GitHub repository to the Radix application.
  • Add Radix SSH deploy key to the GitHub private repository.
  • Set Radix secrets, as well as build and deploy the Radix application.

When the Radix application is available online, it will open in your default browser.

Redeploy

Overwriting, or redeploying, an existing Radix Webviz application is easier. You do

webviz deploy radix ./your_portable_app owner/reponame

and Webviz will:

  • Guide you through Azure CLI login procedure (if not already logged in).
  • Upload the new static resources to the blob storage container.
  • Create a new commit in the GitHub repository https://github.com/owner/reponame.
  • The new commit automatically triggers a new Radix build and deploy.