[CentOS] WARNING: Kernel Errors Present in logwatch

Did you run an enxtended SMART self test on the drive? If two disk
drives are reporting errors, I would start to wonder if the controller
isn't at fault. The SMART self test should help with isolating the
faulty component.

Thanks,
- Ryan

You do realize, that logwatch reports from system logs, and thus might
very well have been reporting the errors from before the disk swap?

Maybe I don't understand the problem...

Matty,

Here are the logs for today. I waited a few days just to be sure (as
Maciej said) that my /var/log/messages was not getting old info.

Dec 8 04:05:27 mail kernel: ide: failed opcode was: unknown
Dec 8 04:05:27 mail kernel: end_request: I/O error, dev hda, sector 6018821
Dec 8 04:05:31 mail kernel: hda: read_intr: status=0x59 { DriveReady
SeekComplete DataRequest Error }
Dec 8 04:05:31 mail kernel: hda: read_intr: error=0x40 {
UncorrectableError }, LBAsect`18827, sector`18821
Dec 8 04:05:31 mail kernel: ide: failed opcode was: unknown
Dec 8 04:05:31 mail kernel: end_request: I/O error, dev hda, sector 6018821
Dec 8 04:05:45 mail kernel: hda: read_intr: status=0x59 { DriveReady
SeekComplete DataRequest Error }
Dec 8 04:05:45 mail kernel: hda: read_intr: error=0x01 {
AddrMarkNotFound }, LBAsectb34661, sectorb34661
Dec 8 04:05:45 mail kernel: ide: failed opcode was: unknown
Dec 8 04:05:48 mail kernel: hda: read_intr: status=0x59 { DriveReady
SeekComplete DataRequest

seems that the controller is bad. so the motherboard is bad.

In an ATA drive, the 'controller' on the motherboard is more like a host
adapter like with SCSI (original ATA (AT Attachment) actually extended the
ISA bus, minus addressing, across to the drive, thus 'Integrated Drive
Electronics'). The 'controller' that is throwing the error is actually on
the drive; more than likely this string of errors is a drive failure. Try
another drive first, then check your cables (cable problems can do this too),
then check your motherboard.