oauth2-cc monitors not recovered after long outage

[fabry006]

📑 I have found these related issues/pull requests

N/A

🛡️ Security Policy

• I agree to have read this project Security Policy

Description

all our monitors configured with oauth2-cc detected an issue with our oAuth2 provider as expected.
The IDP had been down form multiple hours.
As soon as the IDP went back the monitors didn't auto recovered (we let them re-try for the whole weekend).
The only way to fix it was to pause / un-pause all the monitors.
Is it possible that they were trying to re-use an old token instead to ask for a new one?
In our logs we see that Uptime was providing an invalid (or expired) token. Unfortunately the token is redacted from the logs and cannot provide more info

👟 Reproduction steps

I can tell you what happened, haven't tried yet to reproduce it:

• HAve a oauth monitor that works
• shutdown the token provider IDP for 3/4 hours (maybe just longer than the tocke expiration date)
• start again the token provider
• the monitors shouldn't recover
• pause / un-pause the monitor

👀 Expected behavior

The monitors should recover automatically and try to fetch a new token

😓 Actual Behavior

Monitors do not recover. They staid "OFFLINE" for 2 days until we pause / un-pause them

🐻 Uptime-Kuma Version

1.23.13-debian

💻 Operating System and Arch

Docker

🌐 Browser

Chrome

🖥️ Deployment Environment

• Runtime: Docker
• Database: N/A
• Filesystem used to store the database on: local storage
• number of monitors: ~80

📝 Relevant log output

2024-08-26T09:01:35+02:00 [MONITOR] WARN: Monitor #312 'oauth-test': Failing: Request failed with status code 401 | Interval: 120 seconds | Type: http | Down Count: 0 | Resend Interval: 0

2024-08-26T09:14:57+02:00 [RATE-LIMIT] INFO: remaining requests: 60
2024-08-26T09:15:18+02:00 [RATE-LIMIT] INFO: remaining requests: 60
2024-08-26T09:15:35+02:00 [MONITOR] WARN: Monitor #312 'oauth-test': Failing: Request failed with status code 401 | Interval: 120 seconds | Type: http | Down Count: 0 | Resend Interval: 0
2024-08-26T09:19:36+02:00 [MONITOR] WARN: Monitor #312 'oauth-test': Failing: Request failed with status code 401 | Interval: 120 seconds | Type: http | Down Count: 0 | Resend Interval: 0
2024-08-26T09:19:57+02:00 [RATE-LIMIT] INFO: remaining requests: 60
2024-08-26T09:20:18+02:00 [RATE-LIMIT] INFO: remaining requests: 60
2024-08-26T09:21:07+02:00 [MANAGE] INFO: Pause Monitor: 312 User ID: 1
2024-08-26T09:21:10+02:00 [MANAGE] INFO: Resume Monitor: 312 User ID: 1

[CommanderStorm]

Since #3903 we now refresh oauth credentials on 401.
See #4500 for further context on what needs to get done to get said PR released. (literally any help from going through PRs and writing changelogs to writing testcases to writing the needded migration helps)