A question for people using the Silicon Image RAID chip (on a8n-sli for example)

[karmadharma]

I have 2x200GB HDs in RAID1 on the SI chip, and I'm wondering if anybody can help me understand a problem I've been having, as the documentation provided is fairly useless.

Basically the fact is that, when XP shuts down improperly (last time it happened it was due to the IRQ_LESS_NOT_EQUAL in UT2003, since I had forgotten to disable NoExecute) the RAID boots up in a 'weird' state where it works properly *BUT* it's extremely slow (from 100-120MB/sec transfer down to 2-3, and over 30ms of seek).

If you load up the SATARAID5 utility the raid set will be 'yellow' (rather than green) and if you go in the info page it says that the raid is in a 'reduced' state (as opposed to 'online'). I haven't been able to find any documentation anywhere about this, and the only way to fix it seems to be to go in the bios, delete the raid set and recreate a RAID1 set with the same drives: this will re-copy the whole drive and things will be ok and fast again.

I am wondering if there is anything else I can do when this happens, as rebuilding the array every time (besides being scary) takes a few hours, and sometimes I even had to open the case and switch the ports around to let the utility let me do it. Some sort of 'checkdisk' program or something would be ideal.

Thanks in advance for any helpful information you might have.

 

[MeatHead88]

I had no luck with that raid controller either. I switched back to the NVidia...at least on the NVidia I can almost predict when it will lock up on me....

 

[mofrack]

I have the same issue, apparently.

The RAID 1 volume was fine after building it. I didnt even have a BSOD like you did before the array went into reduced mode. I just started up the machine one day to find it like that.

Anyone out there know what the deal is with the SI controller and RAID 1 volumes? Could this be due to the drives somehow? Both drives are physically healthy.

 

[SirAllen]

at least on the NVidia I can almost predict when it will lock up on me....

Hmm, is the NV4 chipset known for locks up? I'm planning on RAIDing from it, and stability and reliability are my top priorities. Hearing things like this scares me.

 

[mofrack]

Please stay on topic. This thread is about RAID 1 problems with the Sil3114 controller on the A8N SLI.

I went into the SIL RAID BIOS to rebuild my RAID 1 array to see if it would return to "online" status instead of "reduced". No joy. My SATARaid util still shows the array'ed drives as yellow instead of green.

Booooooooo!

Should I just copy all this data to an external drive and repartition/reformat the drives?

Is anyone else running a RAID 1 array with the Sil controller on the A8N?? ANY feedback is welcome regarding this specific issue.

 

[mofrack]

Originally posted by: Robbidoo
I've got the same issue - It's because the raid controller is still mirroring the two disks.

If you run the Silicon Image java util, you'll see that both disks are yellow. If you then go Window -> Task Manager, you'll probably see a task labelled 'Restore Redundancy'.

On my 250GB disks this process is taking about 5 hours. The problem is that it seems to do a complete re-mirroring EVERY TIME the machine crashes, or even after I resume from standby.

I'm running the 1.1.0.0 drivers from Asus.

Has anyone found a way around this?

I can't use the nVidia RAID because the controller only recognises the drives every 2nd or 3rd boot - they're both on SATA converters.

So yeah, it turns out the array was rebuilding itself, albeit slowly.

The fact that the array has to rebuild itself afer resuming from sleep is REALLY LAME!! I was confused as to why the array was having to rebuild when my machine wasnt even crashing?

Hope a fix for this comes along soon, not a lot of people using RAID 1 to comment on this...