Hello,
this morning my centos 5 server (2.6.18-53.1.4.el5) failed on the dvd device and disabled dma. The dvd-writer (/dev/hdg) is connected to the following controller as a secondary master. 02:07.0 RAID bus controller: Silicon Image, Inc. PCI0680 Ultra ATA-133 Host Controller (rev 02)
There is also an harddisc as primary master connected on this controller. This harddisc is the only device in the system that I put to sleep with 'hdparm -S180 /dev/hde'.
The error messages are: Jan 24 04:54:21 weidenwinde kernel: hdg: cdrom_pc_intr: The drive appears confused (ireason = 0x01). Trying to recover by ending request. Jan 24 04:59:38 weidenwinde kernel: hdg: cdrom_pc_intr: The drive appears confused (ireason = 0x01). Trying to recover by ending request. Jan 24 05:03:43 weidenwinde kernel: hdg: cdrom_pc_intr: The drive appears confused (ireason = 0x01). Trying to recover by ending request. Jan 24 05:11:34 weidenwinde kernel: hdg: cdrom_pc_intr: The drive appears confused (ireason = 0x01). Trying to recover by ending request. Jan 24 05:19:24 weidenwinde kernel: hdg: cdrom_pc_intr: The drive appears confused (ireason = 0x01). Trying to recover by ending request. Jan 24 05:20:06 weidenwinde kernel: hdg: cdrom_pc_intr: The drive appears confused (ireason = 0x01). Trying to recover by ending request. Jan 24 05:20:06 weidenwinde kernel: hdg: status error: status=0x58 { DriveReady SeekComplete DataRequest } Jan 24 05:20:06 weidenwinde kernel: ide: failed opcode was: unknown Jan 24 05:20:06 weidenwinde kernel: hdg: drive not ready for command Jan 24 05:20:06 weidenwinde kernel: hdg: status timeout: status=0xd0 { Busy } Jan 24 05:20:06 weidenwinde kernel: ide: failed opcode was: unknown Jan 24 05:20:06 weidenwinde kernel: hdg: DMA disabled Jan 24 05:20:06 weidenwinde kernel: hdg: ATAPI reset complete Jan 24 05:30:13 weidenwinde kernel: hdg: cdrom_pc_intr: The drive appears confused (ireason = 0x01). Trying to recover by ending request.
Could this error caused by the hdparm sleep command for /dev/hde?
regards Olaf