S.M.A.R.T error can I recover my RAID

[|*BILLY$CLINT*|]

This past weekend my computer's Intel storage matrix alerted me to an error in drive 2 of my RAID configuration. I ran a test on the hard drive and found out that the error is a S.M.A.R.T electrical error. The drive still works and I know that a S.M.A.R.T errors means that the drive can still function but it's just a matter of time before it fails completely.

I do have a 750 gig seagate replacement drive for the failing one but I am not 100% sure if I can re-build my RAID array on the new drive as I don't know that much about hardware I am afraid. I am also unsure of the RAID configuration of my drives (its either 0 or 1) because I do have a C:/ and a D:/ drive, the D:/ is a windows back up drive and from what I have read in the documentation from Intel I have a RAID 1 but there software says that it's a, "RAID 0(stripping)" which is confusing to say the least.

The computer is working just fine and I have not noticed any errors in fact I am using it to post right now as I copy my most crucial data to my external. Normally in this situation I would re-format but I am using about half of my 1.5 terabytes of space as I do a lot of work from home so this down time would make me suffer.

So any advice on what to do or even letting me know I am screwed would be nice as well. I just really want to know what I should do next so I can be as prepared as possible. The funny thing is I turned my PC on that day to look into adding new partions to my hard drive one for work one for personal...LOL talk about being a day late and a dollar short!

 

[haarg]

It sounds from your post like are currently using 2 750GB drives in RAID 0 mode, which means if either drive fails you will lose everything.

Since you say you are only using about half of the 1.5TB, I would recommend hooking up the replacement drive you have, copying everything to it that you need, then to stop using RAID 0. Switch to just having keeping the drives separate, or possibly using RAID 1 (mirroring) which would leave you with half the space but give you the redundancy to deal with future drive failures.

 

[|*BILLY$CLINT*|]

It sounds from your post like are currently using 2 750GB drives in RAID 0 mode, which means if either drive fails you will lose everything.

Since you say you are only using about half of the 1.5TB, I would recommend hooking up the replacement drive you have, copying everything to it that you need, then to stop using RAID 0. Switch to just having keeping the drives separate, or possibly using RAID 1 (mirroring) which would leave you with half the space but give you the redundancy to deal with future drive failures.

So another question for you, If I install the replacement drive and copy the files to it can i then add it to the RAID when I got to make a new RAID 1 array?

Basically can I make a RAID 1 setup from a drive that had data on it and still keep the data?

 

[haarg]

If you have two drives, one with data you want to keep and the other either empty or able to be wiped, it should be possible to make that into a RAID 1 setup, mirroring the first drive onto the second. The procedure to do that however would depend on what kind of hardware you are using.

 

[Hunter]

you might have to start from scratch as the first drive is already in a RAID array and you might not be able to add it to a new array.

 

[haarg]

For the two drives that are in the RAID 0 array, yes, they will need to be wiped or otherwise overwritten once everything needed is retrieved from them. Once he has a single drive in a working condition (the replacement drive he has) it will likely be possible to use it and the good one of his old drives in a RAID 1 setup without wiping the replacement drive. I couldn't say what the procedure would be though.

 

[Hunter]

you're right with having to have one drive setup as wanted first then its a case of setting up the RAID 1, then the second drive will get formatted and will then mirror the first.

A good way to see if it's already RAID 1 is to unplug the faulty drive and see if it still boots etc.

 

[haarg]

Since he says he has two 750GB drives and 1.5TB of space, it pretty much has to be RAID 0.

 

[Phopojijo]

Yeah...

Raid 0 basically means if any drive fails, you lose the data on both drives.

Why Raid 0 exists is because if you have 2 harddrives -- any read/write done on the drive is distributed amongst all the drives... AND you keep all your space.

Raid 1, total mirroring, allows either drive to fail and everything's great... and while both drives are up, read performance is increased (though write performance sucks).

If you have any data on your drives and you don't backup regularly (like my nightly whole-computer backup with a weekly rotation... which I haven't done for a little bit because my room is getting torn apart for renovations) DO NOT USE RAID 0.

Backup all you can off of the drive and consider it dead because it's just a matter of time before it belches and dies.

And when it does, both drives will go down and you'll need to create a whole new array (or NOT).

 

[rejecht]

* To avoid having to reinstall the system, etc, look into disk imaging utilities. Some claim to offer online imaging. That is, a disk image is made while the OS is running. You can then create a RAID 1 volume, and restore the disk image to it, and have the same system up and running.

* RAID 0 is typically useful for scratch disks; i.e heavy audio/video editing, and other temporary work. It doubles the probability of volume failure, though.

* If you go for a RAID solution, get a hardware RAID controller that has a battery backed up cache. It will protect the system in case of power failure. The pending writes will be held in memory for N hours, typically a couple of days. Also know that any hardware RAID controller will limit you to the manufacturer, or probably the model of the RAID card, since they differ in how they store and load the data on the disks. (The RAID configuration is stored on the disks, so when replacing a faulty controller card, it can read and interpret its configuration from the disks.)

* Don't do a "quick initialization" for RAID volumes, or any new hard disk for that matter (no quick format). Do the full disk check.

* If you can tolerate a day worth of downtime reinstalling your system in the event of disk failure, then ditch RAID altogether. RAID 1 is about uptime.

* Use "on-demand mirroring." That is, use disk imaging software to take snapshots of your system from time to time. You can back up this image to removable storage units that are designed to handle impact/general handling. With that said; many external hard disk solutions are unreliable as a backup solution, unless you handle them with great care and avoid moving them around.

gl

 

[Phopojijo]

A couple of points for the above for total conciseness:

1) It doesn't double the probability of failure... it squares it. ((Well... (1-(1 - probability of failure)^2)))

2) RAID 1 has extra read-access speed boosts... which is useful if you are pulling massive amounts of data from a source you very rarely write to (since writes are DECREASED in speed very minorly due to the overhead of having a RAID controller do a bit of thinking).

That said, the above advantage would only be if you're pulling from database queries or something which is not something a home user would do... and chances are a RAID 5 or 6 would be better anyway.

 

[Zur]

One solution would be to save all your files and use the 750 GB drive to store everything you want to keep in an external enclosure. You can then try reformatting your RAID-0 setup to use only for software and a temporary storage space.

I mean, when you think about it, there are a lot of files you won't touch for a while but want to keep. With increasing drive capacities it makes more and more sense to just have a smallish internal drive and keep everything on an external which will have a much lower probability of failure due to the fact that you only power it up every now and then.

I use one of these myself:

 

[Phopojijo]

Yeah that's what I use...

Rotate between 2 1-Terabyte drives every week -- and backup all my internal drives to them nightly.

Nightly backup with weekly archival.

I especially like the eSata.

 

[|*BILLY$CLINT*|]

Since he says he has two 750GB drives and 1.5TB of space, it pretty much has to be RAID 0.

Well that is what is confusing me as you would think that I had a RAID 0 setup but after reading the following article I am not sure.

For those users who wish to combine the benefits of two RAID levels, matrix RAID is the solution. When using two hard drives, matrix RAID allows RAID 0 and RAID 1 functions to be combined, where critical files can be stored on RAID 1, and RAID 0 can be used for non-critical items such as software. In 2005, the matrix RAID capability was expanded to include RAID 5 and RAID 10. For instance, RAID 5 and RAID 0 can be combined to provide higher performance, capacity, and fault tolerance. A user can edit digital video on a high-performance 4-drive RAID 0 array, then transfer it to RAID 5 for protected storage when completed.

I am 99% sure the RAID matrix is what I am using but all the debug info I got from my computer tells me that I have a RAID 0. So I am going to put it off a few more days and see what else might happen.

 

[Hadmar]

That means Matrix RAID e.g. can take half of the space of HDD1 and half the space of HDD2 and use the two halves as RAID 1 and then it takes the two other halves of both HDDs and use them as RAID 0. If you have the full space of two HDDs available your Matrix RAID is setup to use both HDDs in a RAID 0.

 

[TurdDrive]

this is soo weird

i was just reading about this.

 

[|*BILLY$CLINT*|]

That means Matrix RAID e.g. can take half of the space of HDD1 and half the space of HDD2 and use the two halves as RAID 1 and then it takes the two other halves of both HDDs and use them as RAID 0. If you have the full space of two HDDs available your Matrix RAID is setup to use both HDDs in a RAID 0.

So does that means you have to re-do the RAID strip if you have an error like I do, or can I replace one of the drives and re-build the array? Sorry if this question sounds redundant I am a complete NOOB when it comes to this and I want to make sure that I completely understand what I am doing before I do it.

So with that being said I would like to know your suggestions for what to do next when i re-setup my hard drives(HD). I was thinking of doing a RAID 1 setup but I do know that my HD size will then be limited to 750 gig but I am good with that as I am going to get an other drive for the most crucial back up items.

 

[Hadmar]

So does that means you have to re-do the RAID strip if you have an error like I do, or can I replace one of the drives and re-build the array? Sorry if this question sounds redundant I am a complete NOOB when it comes to this and I want to make sure that I completely understand what I am doing before I do it.

I don't know. In theory there could be an option within the management tool you used to create the RAID to replace a disk by mirroring it first but I kind of doubt it.

Depending on what Matrix RAID uses to distinguish the HDDs it might be possible to just use a tool like dd from a Linux live CD to clone the defective drive to the replacement drive and be done with it. Trying this involves the risk of losing all data though. It's the kind of thing you only try out if you want to know if it works and have a backup of everything you need.

So with that being said I would like to know your suggestions for what to do next when i re-setup my hard drives(HD). I was thinking of doing a RAID 1 setup but I do know that my HD size will then be limited to 750 gig but I am good with that as I am going to get an other drive for the most crucial back up items.

If 750GB is enough for you you can have a RAID 1 or a single disk plus a backup disk.

The RAID 1 should help you if one of the HDDs dies since the other one will have all the data.

A single disk plus a disk for backups helps with things like user errors, ****ed up file systems and partition tables. It also might have all the data you need if the main disk dies. That depends on how often you do backups.

Since you want to get another drive for your important data you could go with a RAID 1 and a backup disk. That's pretty much the best you can do with 3 disks if you really don't want to lose data.

The question really is what are your goals, what do you want to archive with your disks?

 

[rejecht]

I am 99% sure the RAID matrix is what I am using but all the debug info I got from my computer tells me that I have a RAID 0. So I am going to put it off a few more days and see what else might happen.

I don't know the history of your system, but you should have access to the management software. If not, get it from Intel.

http://www.intel.com/support/chipsets/imsm/

 

[ZenPirate]

I have a suggestion. Purchase a copy of spinrite. I've recovered numerous drives with it, and two of them had "smart" error. The errors were gone after repair.