Netraid 2M problem

[Mark Gill]

Hi all,
I have a Netraid 2M, with two 36GB drives running Raid 0. They have been fine for at least six months now. Just yesterday the controller notes a 'NVRAM Mismatch' during a reboot. All docs point to going into 'View/Add/Config' and then selecting either the config from the card or from the drives.

Which is where the problem began! As soon as the message began occuring the machine will hang going into the OS. I have also discovered the Ctrl-M no longer works, both when it says to press it, and when it pauses with the error. The error also implies that it will dump directly into the config tool by pressing any key, and a key press sends the machine into the OS boot, not the config util.

Booting from the Navigator CD rom to get the utils causes a hang during the device scan. So, I downloaded megaconf.exe, put it on a DOS boot disk and ran it from there. I can get into the config tool that way, and one drive is not responding. I'm not sure I believe it - during boot I get '1 logical drive found' with a seemingly simple NVRAM issue. I'm wondering if I'm getting a different response because I'm booting into DOS on an NTFS machine and running the program externally.

I would really like my logical drive back! How can I recover/test the drive by trying all options before I nuke and start over? I could swap Netraid cards, but my other box has a 4M, so it wouldnt be the same. However if there is a chance of saving my data, I'd be game.

*whew* What do you all say? :)

[kris rombauts · Accepted solution]

Hi Mark,

the reason why Netraid Assistant hangs during the SCSI bus scan is because the disk drive would pass the SCSI inquiry command which is basically answering the command "Who are you ?".

The next question from the controller is a Test unit ready which is a kind of "Are you ready?".

Because the drive has a problem of some kind i.e. contamination or handling damage it is confused and it thinks it is coming ready.
The disk replies with "drive not ready but I will be ready soon... ". This is where the hang occurs with NetRaid Assistant, it is mindfully waiting for a promised event which will never happen so it hangs.

The BIOS scan in Express Tools just does an inquiry command which is contained in a ROM on the disk drives controller and for which it can respond without the disk mechanism being ok.

Pls check if the "bad" disk does spin up and check in Express Tools for the disk ID string of the disk. If it is visible but the capacity of the drive is reported as 0 Mbytes you'll have to replace that disk.

This symptom i just described above is also a likely symptom of media contamination. The drive can spin up and pass an inquiry command but when it goes to read from the disk media it's capacity it has a problem due to i.e. contamination. The capacity information is stored on the disk media itself.

When you press CTRL M or use you DOS utility, pls reconfigure the array again by either:

1) choosing the NVram contents as the valid one
(the most probable good one here) and save
it.

2) choose a new config and reconfigure the
array again in raid 0 (using the same
parameters) and save it and restart.

Maybe one of these two helps to get rid of the mismatch, if not one disk will have to be replaced i am afraid.
Since this is a RAID0 and if it turns out your disk is dead, as you know it means a loss of data. Raid 0 should not be a choice if you purchase a RAID controller, it's a waste of money since the RAID controller does not offer you much more then a bit of performance due to it's cache but you don't need a expensive card like a 2M for that.

Once you recovered i strongly advise to migrate the RAID0 to RAID5 by adding one extra disk of the same capacity.


regards, Kris

[Mark Gill]

Hi Kris,

Thanks for your thourough response! It was helpful, but I think the card itself it to blame in my scenario, not the drive. Here's what has happened to me since my first post:

Ctrl-M didnt work for me (still doesnt), and I couldnt boot into my OS for troubleshooting. The only way I could get to the card was to boot to floppy, and run megaconf.exe. That process never asked me the question that everyone said I should be getting, the one that allows me to choose the valid VRAM config info. Because I wasnt getting that option, I was suspicious.

So I pulled all cards from the machine, leaving only the Netraid card in. I researched support.microsoft.com and troubleshooted my way to at least getting the machine to boot back up with the mismatch error still happening. Very long story.

Once I got the OS back up, I installed the Windows version of the Netraid assistant and connected to the card. Again, no message asking me to choose the correct config. The tool reported the same as the DOS based tool, that drive 1 falied and drive 2 was ok. Spin down and spin up commands on both drives worked. Issuing a 'Make online' command to the 'failed' drive brought it back online. Array (and all data) was intact. No problem with either drive!

With the combination of me making the OS tweaks and the Netraid changes I now have a system that is back running as it was before the issue. No loss of data. There seems to have never been a problem with the drive - could there have been a glitch on the card? Surge maybe? My UPS going bad? Dunno.

Frustrating process, to say the least. But your knowledge and willingness to respond is much appreciated. Thanks.