We're seeing that SpiceDB is crashlooping in one of our environments (dev), but only replicas. The of the unhealthy instances mention a "failed SASL auth" database issue and connection timeout. I verifed that the credentials are correct (I can connect using the same creds). I also see this message:

{
  "level": "warn",
  "dispatchReady": false,
  "time": "2023-08-24T16:39:17Z",
  "message": "dispatcher failed readiness checks: found expected state when trying to connect to cluster: TRANSIENT_FAILURE"
}