Something to discuss (potentially with support staff over at Alliance) - ever since MFA has been pushed through, a few issues have been encountered. For the most part, these have been resolved by persisting authentication for as long as possible allowing workflows run to completion. However, if MFA needs to re-authenticate, the datalad api calls will fail with an error message about access and consequently the workflow will fail. ~There is no clear indication when this happens nor is there a push for reauthentication.~ This is usually occurs after the cloud service has been restarted (whether manually or via Alliance support staff).
At the moment, I've been having to occasionally check the workflows and run re-authentication.
Opening this as an issue so it doesn't get lost.
Something to discuss (potentially with support staff over at Alliance) - ever since MFA has been pushed through, a few issues have been encountered. For the most part, these have been resolved by persisting authentication for as long as possible allowing workflows run to completion. However, if MFA needs to re-authenticate, the datalad api calls will fail with an error message about access and consequently the workflow will fail. ~There is no clear indication when this happens nor is there a push for reauthentication.~ This is usually occurs after the cloud service has been restarted (whether manually or via Alliance support staff).
At the moment, I've been having to occasionally check the workflows and run re-authentication.