On 2016-10-09 22:44, Marc Chamberlin wrote:
On 10/9/2016 11:18 AM, John Andersen wrote:
Yes, bad sectors cause large slowdowns, but not high cpu. It should be wait states.
Yeah John, I am kinda thinking you are right, something fishy with the file system or disk drives. Carlo's question seems to imply he is also thinking in that direction. I tried to dd clone the one drive that has reported errors and that failed apparently with a lot of disk read errors. So I need to figure out a way to copy everything from that drive to a new one and perhaps change the file system type of the partitions to a better choice. So first question I have is what file system should I choose and if I want to copy everything from the old drive with a partition using a file system of one type, to a new drive with a partition using a different type of a file system, how best to do so? Doesn't look like dd would be a good choice of a tool to use....
If you want to image, don't use dd, use dd_rescue or ddrescue (they are different but achieve the same thing). For data storage I prefer xfs. I like reiserfs a lot, but I limit its total size. All the reiserfs partitions share the same cpu thread, reiserfs does not scale well. If you want to copy files, use rsync.
Also, I am going to want to preserve the MBR in the process.
The program or the partition layout? Or both? -- Cheers / Saludos, Carlos E. R. (from 13.1 x86_64 "Bottle" at Telcontar)