Hard disk error in Intel Matrix Storage Manager

Question

I have two hard disks, both 160GB, and with Intel Matrix Raid on these two disks, I created two volumes

• Volumne0 Raid1
• Volumne1 Raid0

Now, one of the hard disk shows error in the Intel Matrix Raid manager

Questions:

• Why only Volumne0 marked degraded, but not Volumne1? Suppose the error hard disk also contains data for Volumne1, but its health is now marked as green?

• I've tried to check the hard disk with hddguardian and it seems to be fine

• If I am going to replace the hard disk, but as it also contain the Volumne1 which is Raid0, so the olumne will be also offline when I remove the disk, what would be the best strategy for replacement?

Answer 1

Why only Volumne0 marked degraded, but not Volumne1? Suppose the error hard disk also contains data for Volumne1, but its health is now marked as green?

My hypothesis is that your hard drive went offline for some time.

A potential reason why vol0 is degraded but not vol1 is because vol0 is raid 1: mirroring. If one of the two disks went offline, the data are now unsynced (and that is easy to detect).

But in the case of a Raid0 volume (vol1), there is no parity data and no duplicate of the data. So when the disk is back online, everything is good and life continues. There is no way of knowing if data was lost (maybe except checksums, but I doubt Intel Matrix implements that). At best, the controller could have logged the incident and reported you that data was probably lost. But in raid0, data integrity is not of concern.

I've tried to check the hard disk with hddguardian and it seems to be fine

Yes, if any other hardware was at fault (cables, power, controller), the HD could have been disconnected for some time but be 100% healthy. But it is not to exclude that the HD might have a intermittent issue.

If I am going to replace the hard disk, but as it also contain the Volumne1 which is Raid0, so the olumne will be also offline when I remove the disk, what would be the best strategy for replacement?

If the problem is really the disk, the best strategy for replacement would be to add the new disk, re-sync the array with the new disk and then remove the faulty hard-drive.