On Wed, 10 Oct 2001, Florian Friesdorf wrote: > On Wed, Oct 10, 2001 at 08:31:01AM -0400, Peter Billson wrote: > > > Then if one fileserver was down (even temporarily), then all the other > > > fileservers (all four) would have to queue a message about the data and > > > task and some heartbeat between fileservers could alert it when back up > > > and then make sure that the particular filesystem is properly updated. > > > > > > What do you all think about this? > > > > Sounds exactly like RAID except that the disks are in physically > > different machines. I wonder if you can set up software RAID to use NFS > > mounted drives... hmmmm... may be worth playing with. > > No solution, just a direction: > > The Enhanced Network Block Device Linux Kernel Module > "It makes a remote disk on a different machine act as though it were a > local disk on your machine. It looks like a block device on the local > machine where it's typically going to appear as /dev/nda." > "The intended use is for RAID over the net" > http://www.it.uc3m.es/~ptb/nbd/ > > from the Software-RAID-Howto: > "Linux RAID can work on most block devices. It doesn't matter whether > you use IDE or SCSI devices, or a mixture. Some people > have also used the Network Block Device (NBD) with more or less success."
RAID is definitely not what you want in this situation. Look at CODA which is a massively buffered network filesystem originally designed to work in disconnected operation, managing reentry etc. the relevant kernel support is in the main 2.4 kernel tree these days. striping coda filesystems together may be possible, but I have never played with it. Brian [EMAIL PROTECTED] "You can't depend on your judgement when your imagination is out of focus." -- Mark Twain
