Restoring RAID 0 after drive failure

[idata]

I recently noticed that one of two 250 GB hard drive was going out. System was configured to a RAID 0. Port 0 OK, Port 1 failed.

I replaced the defective drive with a 1 TB. When I went to restart the system, it would not start. I removed the 250 GB drive in Port 0 and replaced it with the 1 TB drive. I did not re-install the working 250 GB drive. I loaded Vista onto the 1 TB drive and got the computer running. I have reinstalled the 250 GB drive in Port 0 and moved the 1 TB to Port 1. Now, I want to include the original working 250 GB drive and restore the system to a RAID O.

I have the following info;

1. Under the RADI drives in the Matrix Storage Manager (MSM), list the 250 GB drive as Port 0.

2. Under the RADI drives in the MSM, the defective drive info has been replaced with the words "Missing Drive".

3. Under Non RAID hard drives in the MSM is listed the 1 TB drive as occupying Port 1.

Installing this drive was something new for me and I obviously missed a step. The instructions on this site are not directed towards my situation. Does anyone have some advice to correct this back to a RAID 0 without having to delete my existing RAID volume and have to re-restore my data?

Thanks,

[idata]

RAID 0 has no redundancy. Kiss goodbye to all your data......

next time please do frequent backup to external storage or use something better, eg RAID 10 or RAID 5....

[VPent]

Hi Micah,

On a Raid0 array, all the data is fragmented and distributed to be written in the disks that are part of the array (in your case 2 disks, the one that is good and the other that is damaged).

So, unfortunately, the data on that array is lost, even that in the BIOS of the motherboard, the old (and now broken) array shows up itself saying that one drive is missing.

By the way, although it is possible to make a Raid0 array with two disks with different characteristics, it is recommended that both disks be "identical", that is, same manufacturer, model and capacity.

To my knowledge and practice, it would be very difficult, if not impossible, to recover the data, by simple method, of the broken disk, put it in a good disk, and reenter this disk on the volume that is mounted on the bios (or MSM), because the bios would not recognize the new disk as the one that is missing (serial # s are different,etc), in addition, to make a new volume, all the data in the disks that will be part of the array, is deleted.

So, i think you should delete the old volume in the bios or MSM and make a new one with 2 identical drives.

By the way, although some say raid0 is a bad thing, my experience with a 2 velociraptors raid0 is great, with it working heavy for almost 3 years without a glitch (but for security i always had a backup).

Hope this helps.

[idata]

Valdir,

Thanks for the useful info. I suspected that RAID would have issue with two drives of different sizes. Had I have known, I would have ran the computer on just the 1TB drive and used the 1 TB external as a backup. Instead, I hated to see 250 GB sitting there doing nothing. Now, I only have a 500GB RAID 0 with 500 GB lost in space.

It sure would be nice if this site had some better instructions for setting up a RAID. Nowhere does it say that mixing sizes is a problem. I could have saved myself a few $$$ and a hell of a lot of anguish and time. I have reloaded the OS about 6 times and I hate the MS Vista Update Center. The updates just never end.

I'm just going to leave it like it is. If the 250 GB drive crokes, I will replace it with a 1 TB and reassemble the RAID 0. 500 BG is probably more than I need now.

I did like the RAID 0 set up. It warned me that the drive was failing and I had enough time to get an external drive. I lost nothing.

Just out of curiosity, you said there was a way to mix drive sizes in a RADI, how do you do it?

Thanks,

Micah

[VPent]

Hi Micah,

When we make a raid0 volume with 2 disks of different sizes, the total size of raid0 array is 2 times the size of the smallest drive and the remaining space of the biggest drive is "lost".

So, if we have a raid0 volume made of one disk with 250 GB and the other with 1000 GB , the total size of array will be 500GB (the raid software will recognize only 250 GB of the 1000 GB disk) and 750 GB of space will be "lost" on the 1000 GB disk.

Also, the raid performance will be limited to 2 times the performance of the slowest disk, that is, the maximum performance of a raid0 with 2 disks with 30 MB/s transfer rate is 60 MB/s, but if the transfer rate of one of the disks is only 20 MB/s, this raid0 array will have a maximum 40 MB/s transfer rate.

And so with the MTBF, the resultant mean time between failures of the array will be the the smallest MTBF/2 of the 2 drives.

Practically, to make a raid0 with 2 disks of different sizes depends how the management software of the raid array will detect the 2 disks: if it reconizes the 2 disks you are done, if it don't recognize the 2 disks as able to form a raid array, you will have to make a partition on the biggest disk of same size of the smallest disk.

Yes i agree that the instructions about the IMSM don't cover all aspects of it, but normally the motherboard's User Guide delves into the details of making a raid array, if you use the soft that cames with it.

regards,

valdir