Skip to content

mselwa/status-page

 
 

Repository files navigation

VSC status page

Repository that drives the VSC status page.

What is it?

  1. incidents: directory containing the incidents.
  2. config: directory containing the configuration file(s) for the verifier and the renderer.
  3. scripts: Python scripts to verify incident files and render the web site.
  4. templates: (HTML) templates to render the website from.
  5. docs: directory that contains the actual website, populated automatically by the Github CI/CD.
  6. requirements.txt: Python environment description.
  7. Dockerfile: docker file that contains all software required to to run the scripts.
  8. .github: CI/CD workflows.
  9. dev_incidents: some incidents useful for testing during development.
  10. example_incidents: some example incident files.

How to use it?

Two workflows have been defined:

  1. a workflow triggered by a push to the main branch, and
  2. a workflow that is executed every 30 minutes.

Both workflows render the site into the docs directory. This directory is exposed by Github Pages and will be linked to https://status.vscentrum.be/

To create an incident:

  1. do a git pull of the main branch;
  2. add your incident description as a YAML file in the incidents directory;
  3. commit your change and push;

You will find some example incidents in the example_incidents/ directory.

The push event triggers rendering the website.

The script that runs every 30 minutes will ensure that an incident with and end date will no longer show up when that date is reached.

Semantics

Incident level:

  1. low: you can expect reduced performance, but functionality is intact. For instance, parts of the system are offline for maintenance.
  2. medium: performance is reduced, and some functionality may not be available. For instance some node types may not be available, or non-critical storage could be impacted.
  3. high: functionality is impacted to a large extent. You may be able to log in, but not to launch jobs, and your data may be temporarily inaccessible.

This is a rough indication, the description of the incident should detail its scope and impact for the user.

Affected: The domain are fairly self-explanatory (eg. tier1_compute, tier1_data, tier1_cloud, tier2_antwerp, tier2_xxx, backbone) and are define in the config.yml file. Note however that a problem with local storage may have impact on other systems. If the storage system that hosts the home and data directories of users at an institute is down, these users will not be able to use infrastructure on other sites either. You may want to lists such issues also for the VSC storage domain.

Planned: Note that the planned field is intended for bookkeeping purposes only. By construction, an incident is planned when its start date is in the future. As soon as the start date is in the past, and the end date still in the future, it is a current event.

Incident naming scheme: To keep it a bit organised, start the incident filename with the site that reported the incident followed by the date of report and in case of multiple incidents on a day, add an index (eg. kul_20221021_01)

Development

The config/dev_config.yml configuration file is intended for development and testing. The render script takes an options --now that can be used to set the "current time". The example incidents in the example_incidents directory can be used with --now '2022-01-20 10:00:00', e.g,.

$ ./scripts/render_site.py  --now '2022-01-20 10:00:00'  --config config/dev_config.yml

Note that one of those example incidents contains an error on purpose.

About

Status page for VSC infrastructure

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 66.8%
  • Python 25.1%
  • HTML 6.8%
  • Other 1.3%