even when using 'mfs' for union remotes, somehow, when one gets too full, it can make the sync fail due to "quota limit reached", in the case of microsoft onedrive. the error doesn't say which remote is causing the error though. it'll just say something like
2026/08/17 23:46:55 ERROR : Cancelling sync due to fatal error: HTTP error 507 (507 Insufficient Storage) returned body: "{"error":{"code":"quotaLimitReached","message":"Quota limit reached"}}"
would it be possible to have the error say which one is causing the issue?
Policy is important as well. If you are using MFS (most free space) then your policy should route writes away from that remote, as it is now full. But quota failures occur during sync if the remote exceeds its quota during the transfer process.
yeah in my experience when MFS still blows up mid-sync with a quota error the free-space check was fine at the start and then one leg filled during the transfer. the error almost never names which remote it was which is annoying.
whats helped me is running rclone about on each union member before the big sync so you can see which one is actually tight. sometimes also --union-cache-time 0 once if the free space looks stale. might try a smaller --max-transfer just to make it fail earlier on the guilty remote.