This is a first for me. I'm dd'ing a 17GB partition (/dev/hdb2) to a
file on a 30GB partition on a second physical drive (/dev/hdd1).
# mkdir /mnt/dest
# mount /dev/hdd1 /mnt/dest
# time dd if=/dev/hdb2 of=/mnt/dest/gyro.rootfs.img
The dd process runs for about 15 minutes or so then I get a kernel oops
printed to the console. klogd probably grabbed it and stuffed it into
syslog as well, but I haven't checked yet.
Instead, I rebooted to runlevel 1 and killed every process I could to
eliminate unexpected interrupts. I also unplugged the usb mouse and the
network cable. It's running again as I type this.
While I wait, I was just wondering if anyone has seen this before with
intense disk i/o like this? Anyone have experience debugging kernel
oopsen?
Before you ask, yes, both disks have dma turned on and hdparm reports
that the destination disk is faster than the source disk. The box is
booting a stock RedHat 9 kernel from an ISO install (on /dev/hda)
(text/plain)