> From [EMAIL PROTECTED] Sun Aug 13 12:00:21 2000
> [snip]
>
> So, the final unanswered question is why did the Mylex controler failed
> that ungracefully if no disk contains dead blocks ?
> My experimental conclusion is that Linux software RAID is even more
> reliable (the two RAID sets handled by Linux sofware had no problem since
> I use the right conservative bus speed), and much more flexible (enabled
> me to check disks individually, and freely select the right RAID 5
> configuration to match existing datas).
> Thanks to Ingo and others for this great work.
>
> Generally, the Mylex PCI RAID controllers take disks offline when certain types
> of unrecoverable errors occur. The driver will log the reason for any disk
> being killed as a console message. Without further information as to precisely
> why the disks were taken offline and whether they all were taken offline
> simultaneously, it's hard to know what happened. Firmware bugs in either the
> controller firmware or disk drives are a plausible reason, as would be a
> problem with the SCSI controller chip on the AcceleRAID, or an electrical
> problem on the SCSI bus.
I have also had this problem once with a Mylex ExtremeRAID 1100. On
a reboot, it placed 3 disks of a 9 disk RAID 5 array offline (2 on
one channel and 1 on another channel). Luckily, I was able to force
the disks back online after wrestling with the less-than-intuitive
Mylex bios interface and the array came back online. The problem
has not happened again since then. The disks are 18gig Seagate
cheetahs with very short LVD cabling and active termination on all
3 channels. A bit disconcerting.
For what it's worth, I have had numerous machines over the last
2 years with software RAID 0 and a couple RAID 5. I've never
had one array fail (knock on wood). If software RAID could gracefully
handle hot plugging dead drives (adding a spare without any human
intervention), I'd probably dispense with hardware RAID completely.
Cheers,
Chris
--
Christopher Mauritz
[EMAIL PROTECTED]
