I've created an mhddfs pool over several drives of varying sizes - one 4TB, two 2TB, and four 1TB. I chose mhddfs after reading several threads on these forums which seemed to indicate it was the better choice, slightly slower performance but more stable (at least until LayerFS). I set the threshold to 100%, hoping to mimic aufs by striping the data more evenly across the drives. However, it appears that the drive is selected based on the number of free bytes, rather than the percentage of the drive that is free, so of course all the writes are done to the single 4TB drive (I transferred my music collection to the pool via ftp and only the 4TB drive was utilized). Is there a way to change this behavior so that the drive is chosen based on percentage? Is this a terrible idea and not something I would want anyway?
Jetzt mitmachen!
Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!