Drives failing naturally at the same time - highly unlikely!
I would think:
- PC got bumped really hard while drives were in use
- other problem (motherboard, power supply, overheating, etc.)
Thanks for replying Jim. I'm confident the PC hasn't been bumped or anything. There has been a noise coming from a fan ( i think) on start up recently. Haven't idenitified which one yet, it's a kind of grinding noise, which goes completely after about 20 secs. All fans seem to be working fine right now, and the CPU is running at <30C. Are there any software tools you know of to test PSU or overheating on the drives, for example? Or do i just need to strip things out and see what happens?
My RAID card shows all drive temperatures in a web browser, but with drives connected to your motherboard there are various different tools to view drive temperatures (speedfan, smartmontools, etc.).
I start to get a bit concerned if my rotating drives are hitting in excess of 43 deg. C. If you are ever seeing 50 deg. C plus, that could be a big concern.
Just reporting back on this. I put two of the supposedly 'failed' drives into an old external G-Raid caddy, and they worked perfectly, passing SMART and chkdsk tests with flying colours. Still had problems with creating the raid 0 onboard though. I bought a PSU tester and checked the PSU unit - seemed fine, then did each SATA power cable one by one. One of them was faulty (it was a 2 way splitter), replaced it and for good measure took out and reconnected all the SATA cables. Et voila, all stable and working beautifully again (so far!) Thanks Jim for pointing this hardware novice in the right direction, and I hope this tale is useful for someone else in the future.
I often experienced this with my Windows 7 machine using the built in software RAID configuration but using a RAID 1 array. This seemed to just be something Windows did from time to time. Being RAID 1 I would just re-add the 'failed' disc and it would be happy again for a number of months. I don't believe the disc ever experienced a failure, you could mount the failed RAID 1 disc as an independent disc and it had the mirror of the data on the first disc. At the end of the day trusting Windows to do your RAID is not a reliable solution and something I would never try with RAID 0. Even a low end software based PCI-E or motherboard based RAID card will prove more reliable.
Actually HDD's often idle 50C to 52C with the current crop of platters. That is not out of range and not an issue. The drives don't start to throttle till over 65C and become an issue at 70C. You normally add around 4C to the temp of an Idle drive for load so even at 52C idle your talking 56C at load most of the time. Failure rate is not tied to temps. It's tied to production and quality of components coupled with mechanical moving parts. HDD's have the highest failure rate out of all computer components for a reason. The old days when drives rarely failed and lasted 5 years easily are long gone.