On Thu, 2005-09-15 at 16:20, Brandon Beattie wrote: > On Thu, Sep 15, 2005 at 04:08:02PM -0400, Tom Lichti wrote: > > But you wouldn't need to shrink the filesystem to replace the disk with > > RAID5, if you use enough disks (minimum 3). Of course there is a trade > > off in terms of available data space, but if what you put on there is > > important to you, it's worth it. At least for me it was. Performance is > > increased as well, since you have more spindles to read and write from > > simultaneously. > > .. You need to re-read what I said. My point was not avoiding a failed > disk problem. My point even exists if you simply want to remove a disk, > for any reason -- You can't unless you're going to put a new disk in its > place, that is the problem/hidden pitfall to XFS, JFS, and Reiser4 with > LVM. I have no desire to get into the raid debate because it's > completely unrelated to the problem I brought up. If you're curious why > I chose not to do raid, e-mail me and I'll explain off-list. > > --Brandon
If you are concerned about failed drives LVM is not a solution. RAID is a solution to handling failed drives. You have an excellent point regarding that fact that you can not shrink certain file systems. IMHO trying to use LVM to provide a solution for handling failed drives or as a method of replacing failed drives is not the right tool for that job. I currently run a 1TB XFS file system for one application, no raid. If/when there is a failure I will lose data. For this application I don't consider that a major problem. :)
_______________________________________________ mythtv-users mailing list [email protected] http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
