SUMMARY: Ultra 5 internal disk timeout and disappear during heavy I/O

I happily report the problem is resolved.

Many people think it is a disk problem, and advised me to run
the read analysis. I did but I did not find any errors.

Some say it is a "BIOS" (openboot) problem. Officially Sun supports
only 2x20GB internal disk drive for Ultra 5, and Sun uses Seagate.
But when Sun supported 1x9GB drive, I already used 2x17GB disk drives.

I resitted the cables, and upgraded the OBP to
"OBP 3.31.0 2001/07/25 20:36". I am so sure which action did the
trick, but I tend to believe it is the later.

I did Oracle import 5 times (each took 2.5 hours), and all of them
were successful.

I hope this info is of use for people who uses Ultra 5 and want to
use large capacity internal disks for testing and development.

Thanks all.

On Sun, Oct 21, 2001 at 12:46:33PM -0400, Michael Wang wrote:
> During certain I/O operations (Oracle import so far), the filesystem
> disappeared, "fsck" says it can not find the device, even the
> "probe-ide" at ok prompt can find the disk. The only way to get the
> disk back is power cycle the box.
> 
> Here is my box configuration:
> 
> machine = Ultra 5
> CPU = 1 x 360MHz
> memory = 320 MB
> OS = Solaris 8 Maintenance Update 5 applied
> Kernel patch level = 108528-10
> 
> Disk vendor = Maxtor
> Size = 40GB
> Model# = 5T040H4 Revision=TAH71DP0
> RPM = 7200 rpm
> ATA = ATA/100
> Master = Yes
> Jumper = factory default
> 
> Here is the error message:
> 
> Oct 15 02:29:19 mercury scsi: [ID 107833 kern.warning] WARNING: /pci@1f,0/pci@1,1/ide@3 (ata0):
> Oct 15 02:29:19 mercury         timeout: abort request, target=2 lun=0
> Oct 15 02:29:19 mercury scsi: [ID 107833 kern.warning] WARNING: /pci@1f,0/pci@1,1/ide@3 (ata0):
> Oct 15 02:29:19 mercury         timeout: abort device, target=2 lun=0
> Oct 15 02:29:19 mercury scsi: [ID 107833 kern.warning] WARNING: /pci@1f,0/pci@1,1/ide@3 (ata0):
> Oct 15 02:29:19 mercury         timeout: reset target, target=2 lun=0
> Oct 15 02:29:19 mercury scsi: [ID 107833 kern.warning] WARNING: /pci@1f,0/pci@1,1/ide@3 (ata0):
> Oct 15 02:29:19 mercury         timeout: reset bus, target=2 lun=0
> Oct 15 02:29:51 mercury uata: [ID 568058 kern.warning] WARNING: Drive not ready before set_features
> Oct 15 02:29:51 mercury dada: [ID 107833 kern.warning] WARNING: /pci@1f,0/pci@1,1/ide@3/dad@2,0 (dad1):
> Oct 15 02:29:51 mercury         transport rejected (-2)
> Oct 15 02:29:51 mercury dada: [ID 107833 kern.warning] WARNING: /pci@1f,0/pci@1,1/ide@3/dad@2,0 (dad1):
> Oct 15 02:29:51 mercury         transport rejected (-2)
> Oct 15 02:29:51 mercury dada: [ID 107833 kern.warning] WARNING: /pci@1f,0/pci@1,1/ide@3/dad@2,0 (dad1):
> Oct 15 02:29:51 mercury         transport rejected (-2)
> Oct 15 02:29:51 mercury dada: [ID 107833 kern.warning] WARNING: /pci@1f,0/pci@1,1/ide@3/dad@2,0 (dad1):
> Oct 15 02:29:51 mercury         transport rejected (-2)
> Oct 15 02:29:51 mercury dada: [ID 107833 kern.warning] WARNING: /pci@1f,0/pci@1,1/ide@3/dad@2,0 (dad1):
> Oct 15 02:29:51 mercury         transport rejected (-2)
> Oct 15 02:29:51 mercury dada: [ID 107833 kern.warning] WARNING: /pci@1f,0/pci@1,1/ide@3/dad@2,0 (dad1):
> Oct 15 02:29:51 mercury         transport rejected (-2)
> Oct 15 02:29:51 mercury ufs_log: [ID 702911 kern.warning] WARNING: Error reading master
> Oct 15 02:29:51 mercury ufs_log: [ID 127457 kern.warning] WARNING: ufs log for /export changed state to Error
> Oct 15 02:29:51 mercury ufs_log: [ID 616219 kern.warning] WARNING: Please umount(1M) /export and run fsck(1M)
> _______________________________________________
> sunmanagers mailing list
> sunmanagers@sunmanagers.org
> http://www.sunmanagers.org/mailman/listinfo/sunmanagers