No, the computer isn't oveclocked in any way. I've started with perfectly working system which was running Vista on IMSM-based RAID5 for 24/7, since over two years. All I did is I've taken one drive out of three and instaled Windows 7 (RTM), but created RAID1 array this time. And the verification error frenzy began. I've tried several fresh installs since then, also after full zero-fill of the drives. Now I'm performing full drive surface tests using SeaTools.
I am little surprised that noone else is experiencing that. Anyway, I've found a kind of workaround. I've once again made both drives non-RAID, then cleared them using Seagate's tool. Then CTRL'd-I at startup and created the RAID1 array. Installed Windows 7 on that array using custom driver (formely "F6") option and providing IMSM "floppy" files from USB stick. However, that step doesn't seem to matter - if I install using Windows built-in drivers and then change to IMSM-supplied drivers, everything behaves the same way. After the system was up and running, installed IMSM. If after this point I performed "Verify and repair", IMSM would report huge amounts (counter does stuck at 65535 after some ~12% of 160GB array) of supposed verification errors. I'm not sure if actual data corruption takes place during that "repairing". What's worse, some smaller amounts of errors would pop up pretty much every time I'd try to run the verification later, even after initial full pass of "Verify and repair". What I've done this time instead, was deliberately degrading the volume by shutting down, pulling one of the drives out and then powering back up in this state. IMSM has reported that the drive is missing and the data integrity is in danger, of course, but otherwise everything seemed to work fine. So I shut down, reconnected the drive and let IMSM do the rebuild. Couriously, after this rebuild, IMSM seems (I've done only few tests so far) happy and does no longer detect any error - every verification pass I've ordered so far completed cleanly. Looks like some serious bug to me. I probably wouldn't experience this kind of problems if I installed the OS on one drive in "RAID-ready" setup and then migrated the data to RAID1, but in this scenario, IMSM wouldn't let me create RAID1/RAID0 mixed arrays on the same set of drives, forcing the array to take whole disk space.
When my Raid1 system would freeze and therefore require a hard restart, MSM would initiate the raid verification process once I logged in. The process showed verification errors maybe 1 out of 5 times. I zeroed the drives before creating the Raid1 volume. It's hard to say whether the verification errors were due to paging space on the disk. Is there a way to pinpoint where on the disk the paging space is allocated and whether it was in that region the verification errors occurred?
I have the same problems like Zbig. I installed Windows 7 x64 on RAID 1 array of two Samsung Spinpoint F3 500 GB hard drives. Just after installation I upgraded Intel Matrix software to 8.9 and ran array verification. I received 65535 verification errors (the same amount to Zbig's result). Both drives were previously tested with HD Tune Pro (full surface scan) with no errors. Write and read benchmarks didn't point to any problems too (116 MB/s avarage read and write - both disks).
Windows 7 is not the first OS I tried to install in such configuration...
One year ago I installed Windows Vista x86 on RAID 1 array of two Samsung Spinpoint F1 320GB (NOT F3 500GB). I did not hit upon an idea to check newly created array (both disks checked with HD Tune too), but after a few month I started to experience problems with explorer freezing after about 20 minutes idle. After the first time array verification started automatically with the result of about 6xxxx verification errors (I dont remember the exact number, but I assume that might be 65535...). The problem returned periodically (always after leaving system idle) until I used disk defragmentation. After that freezing problem stopped, but after leaving system idle I sometimes experienced strange "long folder refreshing" (I took about 30 seconds to refresh any folder until I reboot system). What more: every verification scan after the first showed up dozens of errors and sometimes ECC errors.
My system configuration:
ASUS P5K-Pro (ICH9R controller)
Quad Core Q9550 - no overclocking (now on windows 7, on windows vista I had Core 2 Duo E6600)
The rest is irrevelant, I think.
Hard drives seems not to be guilty since I tried two different configurations (or maybe Samsung is guilty ).
Is anyone know the solution for this problem?
Thanks for reporting all these experiences everyone! Yes ZBig, you are not the only one experiencing these.
I'm experiencing similare. I have two WD 1 terabyte drives in a Mirror RAID(1) configureation, and do full BIOS baseed disc scans on each all the time with no errors. I have Windows 7 (64).
But I constantly get system freezes and slowdowns, and I can tell the disk is trying to re-read data durring these times.
Constantly the system crashes, and when it reboots, Matrix Storage Console Status is: "Verifying and Reparing" and takes hours to complete. It often reports that dozens of "Verification Errors" have been detected and fixed durring this process. What are these and what do I need to fix so they will stop?
I've tried to find some SMART disk utilities to see if any of this is being logged by the drive SMART system, but can't find anything that works with these disks when they are in the motherboard RAID. How do you find if there are any SMART errors, or any other SMART data, when disks are managed by Intell Matrix Storage?
The only solution I can see is to stop using these in RAID(1)? which would totally suck!!
I had a very similar issue on my media center box using a P55 chipset with ICH9R. (Windows 7 x64 Ultimate). There is no overclocking, undervolting, or anything else going on, and the case is very well cooled.
The drives are old 80gb seagates in a RAID1 configuration. The drives were moved from my ICH10R machine and were in good condition (albeit old).
I had to rebuild every few days, and sometimes it would not even finish the rebuild without an error. All drive tests, etc showed nothing wrong, and the the drive that dropped out was not always the same.
I ended up rolling the drivers back to 126.96.36.1999, the same ones that were on my ICH10R machine.
Since then all the problems have stopped, and its been flawless for at least 6 months.
Give those drivers a try, and see if they work for you.
I am having the same problem with Windows 7 x64 with service pack 1. I have the latest version 9.6 installed, but had the same issue with 8.8.
Everything seems to work fine, but when I choose "Verify" or "Verify and Repair" option in the IMSM Console, I'm getting many verification errors. Have been unable to find what if anything can be done about this, and it's significance.
I am experiencing the same problem with verification errors every time I run the IMSM Console. I am running the latest version of the driver 9.6. I am also running Windows 7 x64 Ultimate, with two Seagate 500 gig drives (new) in a RAID 1. Have run Chkdsk and other utilities which all indicate drives have no problems. Like zBig I do not think that anything is wrong with the drives. Does Intel follow these forums, has anyone gotten any technical support on this?
Hi, got the same problems here - running win7 ultimate x64 (fully up to date) on gigabyte GA-EP45-UD3LR (rev 1.0) with ICH10R chipset. Intel driver version 188.8.131.524, latest mb bios (F11, i think). Got two raid 1 arrays setup, one with two Western digital 1TB disks wo media errors and one w two old 300 GB drives, seagate i think and again no media errors. There doesn't seem to be any data corruption at the application level, or at least none that i've detected yet, but i keep getting verification errors on the arrays wo there having been any crashes or malfunctions. The default setting for the motherboard is to do some allegedly safe overclocking, but i've switched all these settings to force normal/safe settings and my cooling system keeps all components well within temperature tolerances.
Do the verification errors put me at risk of data corruption/loss?
I have the same or similiar problem. I have Windows 7 64bit on a Dell XPS9100 using Intel Rapid Storage Technology 10.6.0.1002 with two WD500's in a Raid 1. Almost every day System Mechanic reports drive errors (I don't know where System Mechanic gets this info). When I run verify on IRST it reports verification errors of 2 or more. The first time I ran IRST it reported 57 verification errors. SMART doesn't report any problems. I will start running windows check disk before IRST to see what it finds.