[Violet_Shift]

So, I run a 4x 3TB RAID-5 on the Intel motherboard RAID system for my home data storage needs and it's been serving me well for the past few years.

 

A few weeks ago, it fell into degraded mode, and I unplugged and re-plugged the culprit and moved on after the rebuild, business as usual.

 

Today, the same disk failed again, except this time the RAID utility outright told me that the disk had failed (rather than it had merely fallen out of sync, as I presumed last time).

 

To test whether it was dead, I unplugged it from the array and plugged it into a different board, and lo and behold, the disk shows as uninitialised space.

 

 

Then, wondering what was going on, I put it back into the array, and now it simply says it's fallen out of sync, and is currently rebuilding according to RST.

 

 

 

I've got no idea what's going on. Should I be worried that there's something wrong with this disk? All of my disks are from the same batch (they have sequential serial numbers) so I'm beginning to worry about my data.

 

I'm tempted to just replace the questionable disk, although that costs money that I don't have. Could it just be the connection, in which case maybe a cheap SATA cable would fix my problems?

 

Attached is RST's current status, with the drive that "failed" highlighted.

[Advertisements]

Seeing as how you haven't had an official response from a MOD or techie, here is something I copied/pasted from another forum with someone having a similar issue:



There are 3 steps to checking the integrity of data on a RAID:

1. The ability of all hard drive sectors to adequately store and retrieve data. This is a hardware-level verification. Spinrite or the hard drive manufacturer's disk tools are necessary to make this check. You can use also use Windows utilities like HDTune and/or chkdsk /r but it's best to run a DOS-based utility on the raw drive. Some RAID controllers can do this as a background preventative maintenance task, and call it "Media Patrol" or "Surface Scan".

2. Checking that the RAID volume is, in fact, redundant at all block/stripe levels. This is what the Intel utility's "verify" option does. This, for example, makes sure all data is mirrored to the other drive in a RAID 1, or makes sure that all parity blocks are correct in a RAID 5. This does not check the file system. Some RAID controllers can do this as a background preventative maintenance task, and call it "Redundancy Check".

3. Verifying that the file system structures like the MFT are consistent, without any lost clusters or indexes. Windows chkdsk, Linux fsck, etc. are the tools to do this.




Good luck! 

[Fusionbomb]

Seeing as how you haven't had an official response from a MOD or techie, here is something I copied/pasted from another forum with someone having a similar issue:



There are 3 steps to checking the integrity of data on a RAID:

1. The ability of all hard drive sectors to adequately store and retrieve data. This is a hardware-level verification. Spinrite or the hard drive manufacturer's disk tools are necessary to make this check. You can use also use Windows utilities like HDTune and/or chkdsk /r but it's best to run a DOS-based utility on the raw drive. Some RAID controllers can do this as a background preventative maintenance task, and call it "Media Patrol" or "Surface Scan".

2. Checking that the RAID volume is, in fact, redundant at all block/stripe levels. This is what the Intel utility's "verify" option does. This, for example, makes sure all data is mirrored to the other drive in a RAID 1, or makes sure that all parity blocks are correct in a RAID 5. This does not check the file system. Some RAID controllers can do this as a background preventative maintenance task, and call it "Redundancy Check".

3. Verifying that the file system structures like the MFT are consistent, without any lost clusters or indexes. Windows chkdsk, Linux fsck, etc. are the tools to do this.




Good luck! 

[Violet_Shift]

What I think has happened is that one of my SATA cables is faulty. I've swapped out the cable from the offending drive for a spare, since the drive came up as "failed" again during the previous rebuild.

 

If it happens again I'm going to replace the drive. I hope it doesn't come to that.

 

All of my data on the RAID is intact from my investigations so... I hope it's just a bad cable.

[Fusionbomb]

Yes, hopefully it is just the cable that is bad.  I'm just curious, did you run the chkdsk command?  If so, did it find any errors?

[Violet_Shift]

There's nothing wrong with the file system (chkdsk comes up clean), and that disk just decided it's "failed" again. Same one.

 

Maybe it's a power issue. I'll change out the power cable too, and if it fails again. that leaves the disk itself or the RAID controller as the culprit.

[jdmanchester123]

RAID 5 consists of block-level striping with distributed parity. Unlike in RAID 4, parity information is distributed among the drives. It requires that all drives but one be present to operate. Upon failure of a single drive, subsequent reads can be calculated from the distributed parity such that no data is lost.if problem is fix as it is then Data Recovery can be easily done by a professional only .

[terry1966]

raid 5 your data's fine just as long as only 1 drive fails, that drive has already failed a number of times already whatever the reason, so don't mess about backup your data now then change it for a new one is my advice, other wise yes you do run a serious risk of losing some/all of your data costing a lot of money to get it professionally recovered with no guarantees and they'd need all 4 drives leaving you without a system or needing to buy new drives in the interim.