Jun 13, 2009 1:14 AM
matrix manager say drive failed, seagate drive test software says no it isn't
I asked this in graphics and chipsets about a week ago but no-one answered it. I noticed more RAID related questions in this topic, so I am reposting here instead of bumping the old one.
Intel RAID Controller: Intel(R) ICH8R/ICH9R/ICH10R/DO/PCH SATA RAID Controller On an Asus P6T Mobo using Windows 7 RC1. I have updated the Mobo bios to latest firmware.
I have two volumes defined: one for system disks and one for storage. The system volume is 2 x 640 GB WD drives in RAID1. The storage volume is 4 x 1 TB Seagate Barracuda drives in RAID10. All drives were brand new as of two weeks ago.
- Periodically, the Matrix RAID is telling me that a drive has failed.
- The first time I replaced the drive with a spare and tested the supposed bad hard drive. The drive was good.
- Periodically the Matrix Manager would report that different drives had failed.
- That is, it wasn't necessarily the same drive that was reported as failed each time. It was more or less random.
- When it happened it would bring the system to its knees until I was able to coax the system to shut down and reboot... it could take a half hour for the system to close open apps before shutting down.
- That is, it wasn't necessarily the same drive that was reported as failed each time. It was more or less random.
- I stopped the (re)boot process at the RAID configuration menu (not the board bios)
- I 'removed' the bad drive and then added it back and resumed the boot process
- I did NOT physically remove the drive and replace it.
- I 'removed' the bad drive and then added it back and resumed the boot process
- After Windows started, the Matrix Manager told me it was rebuilding the array which it was.
- The system performance was pretty close to normal until the rebuild completed.
- After this happened about ten times, I went to the Seagate web site and downloaded their diagnostic software.
- This was happening mostly to the Seagate drives, but occasionally to the WD drives.
- This was happening mostly to the Seagate drives, but occasionally to the WD drives.
The Seagate diagnositic software came on an iso file to create a bootable CD with the software on it.
- You need to go into the mobo bios and tell the system to treat the drives like regular single drives again.
- You reboot the computer with the CD in the drive and it boots to FreeDos and then runs the diagnostic software.
- The program is an old school crude dos type graphical interface with support for the mouse.
- The Diagnostic software discovers your drives and allows you to do a number of tests including a 'Short Test' and a 'Long Test'.
- I only tested the Seagate drives since they made up 4 of 6 drives
- good enough sample size to evaluate whether the problem was the drives or the RAID system since the RAID reported failures happened on more or less random drives when it happened.
Drive Test Results
Short Test
- The Short Test checks thing like the SMART status, whether it had ever failed or marked as failed, operating temperatures, etc.
- The Short Test reported all of the Seagate drives to be GOOD. None had ever failed or were marked as bad.
Long Test
- The Long Test, does the short test plus checks all the blocks on the disk and takes a couple of hours to run.
- The Long Test reported the drives as GOOD.... no failures, errors, etc.
OK People, What's The Scoop?
Why would the Intel Matrix Manager etc. continuously and ERRONEOUSLY tell me my drives are failing or failed and bring my system performance down to its knees, and be absolutely a pain in ths _ss? The drives are good.
Anyone tell me how to fix this other than return the Mobo or buy a real RAID card (which I am 3ware 9559SXU-8PL... but I still want to know what is wrong with my Mobo). BTW, now I know why Linux fanboys refer to cards like the 3ware card I am buying as 'real' (the one I am buying has an actual controller risc chip and its own onboard memory (128MB)), and this Intel RAID as FAKERAID. This experience almost has me in the Linux fanboy club as I am very angry about this.
Any help is appreciated.