f8b9e4d780
[sentry](https://twenty-v7.sentry.io/issues/6545999328/?environment=prod&project=4507072499810304&query=is%3Aunresolved%20issue.priority%3A%5Bhigh%2C%20medium%5D&referrer=issue-stream&sort=date&stream_index=2) Our workers have failing jobs because they are concurrently creating datasources on different pods While the datasources are independent, they read from the same redis cache, with the same `RolesPermissionsOngoingCachingLock` value in redis. <img width="852" alt="Capture d’écran 2025-04-18 à 18 00 20" src="https://github.com/user-attachments/assets/42ce8479-acc1-462b-af4c-b547cc2bb0b8" /> As a consequence they can fail to create a datasource: the first datasource computes the permissions and sets ongoingCachingLock to true, then the second arrives, there are still no available permissions in the cache, but because of the lock it early returns without permissions and fails. This behavior goes unnoticed on the product and helps performances, but is annoying for workers as jobs are failing. Let's remove the cache lock when creating the datasource !