My RAID 1 mirror worked fine until a couple days ago, when it was triggered to start a rebuild. It seems to now be continuously rebuilding, in that I saw it go to 95% done, and when I looked a few hours later, it was back to 5% done. I assume one of the two underlying hard disks is flaky and should be replaced -- but which one?? The ROM BIOS control-I merely says the volume is rebuilding, and Intel Rapid Storage Technology likewise says the volume is rebuilding, and the detailed report of the two underlying drives is "normal". In sum, I see no clue from either ROM BIOS or IRST which hard drive to replace.
This is slightly dated equipment and software. The mobo is DX58SO, and the BIOS isn't the latest -- I had to stop updating because Intel's updates were breaking more things than they fixed! IRST is also not the latest: 22.214.171.1244. I figure it's risky to try to update it at this point.
What's my simplest pathway? I could disconnect one of the two drives and boot it and see if the system functions properly; if it does, then I can assume I've found the bad drive and replace it and let it rebuild again. Or if the system doesn't function with the one drive disconnected, I can disconnect the other drive instead, and see if that works any better. But is there something I can do that doesn't involve guesswork? Something that will report, for example, which underlying disk is being written? I don't want to make a mess here; I just want to make a surgical change.
I am sorry to know about your issue.
I believe it is possible to test the drives by connecting one of them as a secondary drive in a different SATA controller (in this case Marvell's eSATA) or a different computer and run the diagnostic utility provided by the manufacturer. When done, you can repeat the process with the other drive, the idea is to test them as single drive and not as a RAID volume since the diagnostic tool probably will only be able to run the test that way.
It is possible you will need to break the RAID volume, but being this a mirror the data will still be accessible and you should be able to use the same computer and controller; then, run the test and rebuild from the healthy drive.
You may also try using different cables or ports in case one of them is defective. We can leave a BIOS or motherboard issue as last resort.
I hope this helps.
Thanks, that worked. I removed the drives and tested in a docking station, and one of them showed a problem. I replaced it and the rebuild went through okay.
It occurred to me afterwards that maybe the Intel SSD Toolkit might have told me something. I just installed the Intel SSD Toolkit, and it reports SMART data for the hard drives underlying a RAID array. This is on a system with no SSD. So it seems to me the SSD Toolkit is actually a generally useful tool, although I haven't seen it promoted as such.
I'm having the same problem as Dogwood but, sadly, I'm in a situation where I can't test them as single drive. So I would like to know if there is any other way to find which drive is being rebuilt ? (Logically if Intel RST starts rebuilbing a volume is because it found a problem in one of the drives, so it must have that information saved somewhere or it wouldn't know which drive needs to be rebuilt).