What I mean is that when I have the old hard drive in the RAID (4 hard disks that look like two disks. One the 2 "disks" is a mirror of the data for redundancy) I can get it into Windows and it works just fine. But, if I install the new hard drive and try to get it to rebuild the RAID like it should, it just fails and will not even touch Windows.
So right now, I am running the Windows setup again to repair the install which is something that I just have to do when going from the new disk back to the old one to make it all work right. Then it will "rebuild" it's self for about 10 hours (I know, screams of fun right?).
Now I am sure some of you are going, well if it works with the old drive, just use that. Well, I sure as hell would if I thought it would stay up. Hard drives do not just drop out of the RAID for no reason, there must be a reason. It’s either the drive its self is bad, or the controller card is bad.
Well I ruled out the controller card by replacing it with the same model (we have many of the same exact model for that purpose). So that only leaves me with one option, and that is that the drive is bad. That is the theory that I am working with since it’s the one that makes the best sense.
Here is the other thing, when I can get the server back into Windows, all the data is there just like it should. Everything works, including the SQL Server Software and our Dynamics GP software. So, I know that the data on the rest of the drives is good, just that one drive is still in limbo.
So what makes this really fucking weird is that we have only dropped one drive (we can drop two before data lost, in theory) So what should happen is when there is a RAID event, it will drop the drive in question that it has issues with. You can re-add it by rebooting the computer (why it can not just re-add it on the spot, with out a reboot, IDK). So when you take the computer down for a reboot, you have the option to just flat out replace the drive or let it try and rebuild with the same drive.
Now see this is where my problems start. Even with the old drive in, it would not reboot into Windows. All I would get it crappy errors like Windows does not exist. At that point I was about to reformat the whole thing, but I still had theory’s I wanted to try. Since our SQL Server Software and Dynamics GP are running on backup server right now, I do have some time to try and get it right, or at least avoid a rebuild on a server (cus there is nothing saying that will work too).
Now I was able to get it back into Windows the first time by doing a side-by-side install of Windows (and then it would let me use the old install for what ever freaking crazy reason and the new install would have errors out the ass). So when I tried that a second time after trying it with the new hard drive and not getting anywhere, it did not work.
But what did work was a "repair" of the current Windows install (the one I want). So I did that, and it worked. So right now I am doing that for a third time after trying the same thing with the new hard drive and it not working. So I am not sure what in hell is going on here.
I have a theory that the MBR (Master Boot Record) is getting shot each time I try and change the drive, but if that was completely true, then I should be able to recover it with the "repair" of the windows install when I did it with the new drive. I also do need to do some research and see where the RAID keeps the MBR (drive wise). I would have thought each of the 4 drives would have a copy and it could use any copy it needed (since they would all be in sync). But if it keeps the MBR only on "Drive 0", which happens to be the drive that dropped out of the RAID, then that might be the whole cause of this.
But if that was true, then in theory, the "repair" of Windows with the new drive in should work too. Now there is the chance that the new drive is bad, which I will test tomorrow (I've got my ways). Also the thing about the new drive is its not even close to the old drive. The only thing the same about them is that they are both made by Maxtor.
But the new Drive is SATA/300 where as the other 3 drives are all SATA/150 and the RAID controller card is SATA/150, but what the dumb-dumb at CompUSA told me, SATA/300 is backwards compatible to SATA/150, so in theory it should work. The other thing about this drive is that its 320gb and not 250gb. Now the reason why I am really not too worried about that is that in RAID you can not replace a drive with a drive that is smaller then the original drive.
So that is that. Tomorrow I will have to do my research and see if SATA/300 is really backwards compatible with SATA/150 and even if that is true, if this control card would know what to do with it (the controller card does see the drive and know what model it is, but that does not mean much). There are other options I am looking at.
Normally I would just grab a drive out of the backup server (since it has the same drives and configuration of the main server, but since that is in use, I am not about to forcibly cause a RAID Event there. That would just be fucking stupid. I have tried to buy the same drive that we already have, but they are not made anymore (go figure right?). Now I am working on some leads of some places that might still have this drive in stock (not any near here).
So ya, that is what’s going on with that. The other fun thing is my hours. As of my writing this, I have 37 hours for the week in 3 days. Monday I worked 16 hours. Tuesday I worked 10 hours and Wednesday I worked 11 hours. Now normally 37 hours is about where I am on a Friday at the end of the day, which means in theory I could take the next two days off, but I've got a server to fix, so I doubt that will happen.
Well that is all for now. You are now updated on the crap in my life. Somehow this is not how I thought I would spend my spring break. I really think I will take a day or two off (from work, not school) next week if I can get this dumb thing working. Also sorry if this entry seems a little random, but I did freaking write this at 2am after working 11 hours, so ya.