Re: Errors writing to eSATA hard drive
- From: buck <buck@xxxxxxxxxxx>
- Date: Tue, 18 Mar 2008 22:36:28 -0700
On 19 Mar 2008 00:38:16 GMT, Roger Blake <rogblake10@xxxxxxxxxxx>
wrote:
I'm trying to use a Seagate 750GB external eSATA drive for backup
on a Mandriva Linux 2006 system that is running a freshly-built 2.6.24.3
generic kernel. The controller is Silicon Image 3512 based, and
the sata_sil kernel module is being used for it.
What is happening is that when copying large amounts of data to the drive,
the following set of diagnostic messages appear every few minutes in
/var/log/messages:
------------------------------------------------------------------------------
Mar 16 12:12:55 server1 kernel: ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2
Mar 16 12:12:55 server1 kernel: ata1.00: BMDMA2 stat 0x86c1001
Mar 16 12:12:55 server1 kernel: ata1.00: cmd 35/00:d0:27:6b:4a/00:01:06:00:00/e0 tag 0 dma 237568 out
Mar 16 12:12:55 server1 kernel: res 51/84:00:f6:6c:4a/00:00:06:00:00/e0 Emask 0x10 (ATA bus error)
Mar 16 12:12:55 server1 kernel: ata1.00: status: { DRDY ERR }
Mar 16 12:12:55 server1 kernel: ata1.00: error: { ICRC ABRT }
Mar 16 12:12:55 server1 kernel: ata1: soft resetting link
Mar 16 12:12:55 server1 kernel: ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
Mar 16 12:12:55 server1 kernel: ata1.00: configured for UDMA/33
Mar 16 12:12:55 server1 kernel: ata1: EH complete
Mar 16 12:12:55 server1 kernel: sd 1:0:0:0: [sdb] 1465149168 512-byte hardware sectors (750156 MB)
Mar 16 12:12:55 server1 kernel: sd 1:0:0:0: [sdb] Write Protect is off
Mar 16 12:12:55 server1 kernel: sd 1:0:0:0: [sdb] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
------------------------------------------------------------------------------
(Despite the above, the files do seem to get written to the drive without
corruption -- at least the ones I've spot-checked -- and the ext3 filesystem
remains intact!)
The drive works OK if Windoze is booted up (it's a dual-boot system), so it
does not appear to be a hardware problem. So far internet searches have
turned up others having this kind of trouble, but no solution yet. Despite
the fact that at the moment there seems to be no corruption, having these
many errors is not very comforting.
Thanks in advance for any ideas in solving this!
Try appending two parameters to your lilo boot. The computer on which
I did this had a power supply fail, so I can't give you the exact
append line, but one was BIOSIRQ and the other APICsomething. I was
about to test each individually to see if only one would resolve the
problem when the PS died...
If I recall correctly, I found these in the kernel source tree
documentation.
Assigning IRQs in BIOS was a waste of time. The kernel decided to use
what it wanted.
--
buck
.
- Prev by Date: Re: External disk doesn't boot. Install GRUB on USB key?
- Next by Date: 54Mbps/PCMCIA wlan card for an X terminal?
- Previous by thread: External disk doesn't boot. Install GRUB on USB key?
- Next by thread: Re: Errors writing to eSATA hard drive
- Index(es):
Relevant Pages
|
