Non-System disk on reboot - Linux/ DL 585

[facnj]

We run weekly reboots on a number of our DL 585 G1. Occassionally some servers didn't come up on reboot, it showed
"Non-System disk or disk error
replace and strike any key when ready"

then went to PXE boot.

At that point all we needed to do was to go to to ILO to power cycle the box, it would then come back up. The problem didn't repeat itself every week. It happens about once in 3 months, but because we have a number of servers, then it would be like a problem every week.

The servers run SLES 9, the boot order is
a. CD
b. flooy
c. disks
d. NIC

Has anyone seen this before? Is it possible to be a corrupted MBR? If so, how can I fix it?

thank you in advance

[Matti_Kurkela]

We've had something similar on an old ML370 G3.

In our case, the SmartArray controller apparently went to sleep and did not respond properly at reboot. Power cycling fixed it.

I noticed the SmartArray firmware was quite old and upgraded it to current level. Since then, the problem has not re-occurred.

Your problem does not seem like a corrupted MBR: if it was that, how would it un-corrupt itself at power-cycle?

I'd recommend checking the the BIOS and SmartArray firmware levels, reading the revision histories of the current firmwares and upgrading if the newer firmware has fixes that might be applicable to the situation.

MK

[Steven E. Protter]

Shalom,

I have seen it.

More likely the firmware on the server needs update. Perhaps the hot swap disk needs to be pulled and inserted firmly.

Battery on the raid controller if installed may be problem.

SEP

[facnj]

Thanks for the response and pointed to the potential hardware or raid controller problem.

However, when the problem first surfaced about a year ago, I had already upgraded all the firmware.
The BIOS firmware I have is 02/14/2007 and from what I checked it is the latest available on the HP web site.

The controller firmware I have is 2.66, the latest version HP web site for storage firmware is 2.76, however the fixes do not address the problem, but I am going to upgrade it anyway.

I am considering to get our hardware maintenance vendor to replace the system board, but I like to avoid the overkill is possible. But it is difficult to go back to the clients a few times to upgrade the firmware as we have already done it twice.

Probably I will get them to replace the battery first.

[Gokul Chandola]

Hi,
Yes first you have to replace system battary then BIAS updation.
Most probably you will get solution, because there is issue with finding bias setting.

Regards,
Gokul Chandola

[facnj]

I thought I will be replacing the battery on the controller, why the system battery?

I think you meant BIOS updating, but why do I need to update the BIOS? and what do I need to update?

thanks