pelasgian
2005-03-05, 12:13 PM CST
My external hard drive gives "rejected I/O to dead scsi device" after being stressed for a while (with something like mkfs.ext3 /dev/sda1 -c -c). As a result, the drive becomes inaccessible, and returns only after power cycling.
The details of the system are as follows:
asus mobo with P3 sporting a via apollo 133a pro (and disabled onboard 1.1 usb)
usb 2.0 controller with via vt6202 chipset with two ports
usb 2.0 hdd enclosure "icy box" with a western digital 80GB JB series
(The chip of the enclosure bears the markings:
CY7C68300A-56PVC
0A PHI 0426
CYP 634016
The uhci interface is connected to an iiyama HA202DT vision master pro 512 usb 1.1 4 port hub driving an optical intellimouse and a HP 1300 laserjet (1st port of the vt6202 pci controller card)
The ehci interface is connected directly to the back of the "icy box" (2nd port of the vt6202 pci controller card)
The system is a fedora core 2 with some synaptic upgrades from fedora core 3 which sports a 2.6.8-1.521 kernel with nvidia modules.
I was accessing the drive in "small bursts" i.e. write 3-4 gigs, then unmount, power cycle, lock-unlock the door, shout out of the window, stay with one foot against the wall et.c. and this seems to work , however, I'd rather not get used to this.
I've noticed in google that many people mention a similar problem, with various brands of gear and that these problems appear after kernel version 2.6.7?
I'd appreciate any help (If you want to send donations, I'll pass you a bank account number :D )
The details of the system are as follows:
asus mobo with P3 sporting a via apollo 133a pro (and disabled onboard 1.1 usb)
usb 2.0 controller with via vt6202 chipset with two ports
usb 2.0 hdd enclosure "icy box" with a western digital 80GB JB series
(The chip of the enclosure bears the markings:
CY7C68300A-56PVC
0A PHI 0426
CYP 634016
The uhci interface is connected to an iiyama HA202DT vision master pro 512 usb 1.1 4 port hub driving an optical intellimouse and a HP 1300 laserjet (1st port of the vt6202 pci controller card)
The ehci interface is connected directly to the back of the "icy box" (2nd port of the vt6202 pci controller card)
The system is a fedora core 2 with some synaptic upgrades from fedora core 3 which sports a 2.6.8-1.521 kernel with nvidia modules.
I was accessing the drive in "small bursts" i.e. write 3-4 gigs, then unmount, power cycle, lock-unlock the door, shout out of the window, stay with one foot against the wall et.c. and this seems to work , however, I'd rather not get used to this.
I've noticed in google that many people mention a similar problem, with various brands of gear and that these problems appear after kernel version 2.6.7?
I'd appreciate any help (If you want to send donations, I'll pass you a bank account number :D )