View Full Version : Burly 8 bay Raid 0 problem?

04-03-2008, 01:39 PM

Recently our Burly enclosure with 8 WD 750gig drives setup as software RAID 0, has been doing something strange. Every so often, I hear a click and some of the drive status lights on the front panel go out. My OS will freeze temporarily until the drives start up again. It has happened several times and I cannot determine which drive might be the culprit. Any suggestions as to how to determine what is going on? I cannot recreate the problem at will. It just happens randomly. I am running a MacPro with 10.5.2 and Softraid 3.6.6.



04-03-2008, 02:05 PM
Does SoftRaid report any I/O errors when view details about the drives after an event?

Try running DU on each drive to verify?

04-03-2008, 02:23 PM
What host card, what computer, what firmware is on the card? Exactly what model drives? (server drives can do some really strange things, desktop drives are much more predictable) Making sure any drivers or firmware is up to date can be an important part of this as well.

I have never seen that behavior from the drives before, so we need to figure out why and how it is possible. The more we know of the setup the more likely we figure out what is happening. FYI, we seldom do much testing in SoftRAID, so if there is anything in particular from that direction we may need to talk with the inestimable Mark James for help.


04-03-2008, 02:47 PM
Thanks for the quick replies.

There are no I/O errors in softraid.

As for System setup:

MacPro 3.0ghz 8gig ram os 10.5.2
Caldigit CDFASTA-4e PCIe (not sure about firmware ver.)
8-Seagate 750gig Barracuda 7200.10 (ST3750640AS)

I'll double check firmware and drivers, but I have 2 of these towers running off the same host card. Only one tower exhibits these problems. i will check DU now on each drive.


04-03-2008, 03:16 PM
Hmm, sounds all correct.

Drives should be fine, probably all the drives have 3AAK firmware on them.

If you look in your System Profiler, under Software in the Extensions list you should see an extension called 'SiliconImage3124u.kext' and it should show version 2.0.3.

WHen having a problem with a RAID0 array it may become necessary to break the array down and zero each drive as a test. That way we find out if any one drive is causing a problem. I don't know any other way to determine if you have a single failing drive.