This is a bit of a newbie question and hope this is appropriate for this list, so apologies in advance. We are hoping (if funding is approved) to get a new server to replace an old server currently in use. The server is primarily an NIS/NFS server supporting a software development computing center. The current server has kind of a hodge-podge of IDE and SCSI disks (no RAID) and a SCSI multi-tape backup device, using amanda as the backup application. A previous sys admin set up the user disk partitions on this server to be around 20 GB to match a 20 GB holding partition for amanda. The new server is planned to have a RAID 5 array with 4 disks plus 1 spare. The software base for this would be Linux (Debian woody) using a recent 2.4.x kernel. We plan to use the software RAID facilities in Linux, not a hardware RAID controller. My understanding of RAID is that this would result in a single, large device -- /dev/md0. This is great and I plan put all the user accounts on a filesystem on /dev/md0 and NFS serve those to all the client machines (mostly Linux, some Sun) we have. The thing I am confused about is how this will affect our amanda backup strategy. The array size will be a little over 700 GB. To follow our current amanda setup it seems that the RAID device should be split into multiple partitions with one dedicated as a holding partition for amanda. But I'm not sure this is possible or desireable with a RAID array. Any ideas, comments? Thanks... - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
