mattermost/server/channels/store/retrylayer
Jesse Hallam 6ef73af2cc
MM-65960: Avoid replica race lag when accessing TelemetryID (#34586)
* avoid replica race lag when remembering ServerID

In an HA environment, with a master and read replica, querying the server id from the store runs the risk of returning a value saved to master but not yet replicated. Avoid this by using the telemetry service value directly when available.

Fixes: MM-65960

* Add Get(ByName)WithContext

* explicitly use master for ServerId

* mock GetByNameWithContext

* more mocking

* more mocks
2025-12-11 17:34:36 -04:00
..
retrylayer.go MM-65960: Avoid replica race lag when accessing TelemetryID (#34586) 2025-12-11 17:34:36 -04:00
retrylayer_test.go [MM-61758] Burn on read feature (#34703) 2025-12-11 07:59:50 +01:00