revive refresh token

[fschade]

Description

This pr adds the ability to try to refresh the refresh token one last time before sending the user to the "log out" view.
This scenario only happens in rare cases where the browser sits in the background and the native browser throttling could happen.

Please read https://developer.chrome.com/blog/timer-throttling-in-chrome-88/ for a more detailed explanation.

The problematic part is oidc-auth-ts which uses an internal timer to organize all tasks, the timer then uses setInterval which gets throttled. If this happens the timer gets out of sync and the accessTokenExpired event gets emitted.

In those cases the code now tries to refresh the token one more time before it sends the user to the error view. Please note that the oidc server needs to be setup correct and the access token lifespan should't be too short.

Related Issue

• none

Motivation and Context

be able to keep the web intact if it runs in the background.

How Has This Been Tested?

• manual setup in docker compose with ocis and keycloak

Screenshots (if appropriate):

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)
• Technical debt
• Tests

Checklist:

• Code changes
• Unit tests added
• Acceptance tests added
• Documentation ticket raised:

Open tasks:

• lets discuss if this is something to hit the master or should go into experimental quarantine.

[sonarcloud]

SonarCloud Quality Gate failed.   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

0.0% Coverage
0.0% Duplication

[dschmidt]

I'm ok with this in oC Web, but do you think there would be a chance to fix this inside the library? Then we should at least open a ticket

[fschade]

I'm ok with this in oC Web, but do you think there would be a chance to fix this inside the library? Then we should at least open a ticket

can't be changed inside the lib, the lib is doing the correct thing here, the library has zero knowledge about the implementing spa app and how it should handle those 'rare' cases.