Hi, I am looking for a redundant cluster filesystem which i would like to be able to setup with three servers running linux. The servers will be using normal disks with hardware raid 5. If one of the three servers fails i would like one of the other two to take over all file transfers. Does anyone know of a cluster filesystem that can meet these requirements. Correct me if i am wrong but the lustre filesystem is not redundant or am i mistaken? I would have to use DRBD to make two servers redundant. AFS can only provide read only copies of data if a node fails. GFS only runs on shared storage. Do i have to look at a commercial product like HP-SFS to meet my needs or is there a way round this. How do other people solve there growing storage problems without spending vast amounts of money on expensive comercial solutions. Any help or advice would be greatly appreciated. Regards Simon