How to fit larger disk drives into your Raid 1 system

 When you first build a Raid setup you usually use the latest largest drives you can afford.

At this point can I remind you of the nomenclature. In computer terminology a "K" is actually 1,024 and therefore one MByte is actually 1,024,000 bytes.
From here on "M" and "T" are used very loosely.

My current system used a pair of 320GByte hard drives, then when I had a spot of bother with one of them, I fitted a pair of 500Gbyte versions.
This wasn't particularly planned. The 320 GByte drives had been fine until one day a drive failed and the only new replacement I had was marked 500Gbyte.
I'd already discovered a bit about swapping drives because a couple of my customers had already had a problem.

I supplied several Raid 1 setups a few years ago, using Seagate Barracuda drives which had faulty firmware that meant all the drives would eventually fail catastrophically.
The idea with a Raid 1 system is that a single drive can fail leaving you with a working computer.

If this happens a warning message pops up and the faulty drive should be replaced as soon as it's convenient.

Raid 1 which is "mirroring" or a method of having 100% backup which is maintained invisibly to the user, is a really good idea for someone running a business.

In fact it should really be a requirement.

Normally a hard drive will fail in a random fashion when something unexpected goes wrong, like a bearing fails or a chance hardware weakness; perhaps when a badly welded connection within a chip falls off.

Usually hard drive manufacturers will quote an MTBF or Mean Time Before Failure for their product. This often represents something like 10 or 20 years or more of continuous operation.
In the case of the Seagate drives the firmware problem meant that the drives would definitely fail within a year.
Anyway, to cut a long story short, I had some experience in managing Raid 1 setups.

The first problem one has when it comes to fitting a new drive to replace a failed drive is that the new one must have the same, or greater capacity as the existing one. This is not straightforward as the marked capacity, say 500GByte, is too approximate a figure and the manufacturers specifications should be consulted before purchase to ensure the number of available bytes is either precisely the same, or larger.

Now that I understand a little more about the design of Raid systems, I could get round the problem, but it's messy. Read on and you can probably figure out how to do it.

I won't go into the procedure of constructing a Raid 1 setup because this article intends to explain how to change an existing setup.

In the example below I'm fitting a pair of 2TByte SATA drives in place of a pair of 500GByte SATA drives.

The first step is to buy two identical hard drives. Choose a model with the largest cache as this will usually be fastest.
Next ensure your computer is running the latest version of Intel Matrix Storage Manager (IMSM). I updated mine from version 5 to version 8. I guess there are lots of other pieces of software that will do the job, but I'm familiar with the Intel version.

Open IMSM and note the details of the old drives and their physical connections. This is important as you must use this information later in the procedure.

Make a note of serial numbers and draw a sketch of the mechanical position of the drives, just in case you forget later.
I took the trouble to do the data check offered by IMSM. This is long winded but a good idea.

Turn off the computer and remove one drive, noting which connector it's using.
Physically fit one new drive in it's place and connect the data and power cables.

Turn on the computer.
A warning message will appear during boot up that Raid is degraded.
Open IMSM and you will see three drives listed. One will be marked as missing. That's the one you removed.
Select the new, larger drive and choose "Rebuild to this drive". Data migration will then take place and this will take maybe two hours or more.
Once the migration is complete you'll have a working Raid setup having exactly the same capacity as before.

In fact you'll have difficulty seeing anything other than 500MBytes capacity for the new drive if that was the capacity of the old one. Its other 1.5 TBytes will be hidden from view.

Next, you should switch off and remove the second smaller drive and physically fit the new larger drive, but do not fit its data and power cables at this point.

Booting up the computer will result in another "degraded" message.
The new hard drive capacity will be indicated at boot up correctly, but within Windows it will still be 500GByte.
The next step is to use a partition manager to increase the size of the main partition to fill all the unused space.
I must admit that although I had no trouble changing from 320 to 500GByte, changing from 500 to 2000GByte (ie 2TByte) turned out to be difficult.

The first step is to remove the large drive from its Raid 1 environment using the "Reset disks to non-Raid" option.

During this process there will be a warning message. It will tell you that all your data will be lost. It's (yet another) product of lazy programmers. The message should include the information that data will not be lost if Raid 1 is your setup.

Go ahead and reset the disks.
This will allow you to see the full disk capacity and you can then attempt to increase the partition size. I say "attempt" because two of three methods I tried didn't work.

Previously I'd used Partition Magic. This time I used Acronis Disk Director. Everything seemed to be in place, but the program needed a computer reboot to fix the new partition and a reboot didn't happen. A forced reboot didn't have any effect either.
I then tried Spotmau Partition Genius, but their genius seemed not to have passed his GCSEs and the old partition steadfastly remained at 500GByte.

Why neither of these programs worked is a mystery. One suggested turning off the anti-virus, but to no avail. Any ideas anyone?

The program that worked was Norton Partition Magic. I think that this was the one I'd used a few years back, but then it hadn't been owned by Symantec.
It does come up with a warning message that you'll be going over the 1,024 cylinder limit and you may not be able to boot up. If you can't boot up because of a hardware issue just give up and put back your original drives.

It worked OK for me however and I was rewarded with a full 2TByte partition (excluding a little, I think maybe 8MByte for Windows).
At this point you will have just an ordinary computer with a single hard drive, so the final step is to switch off and plug in the second new hard drive.

After a reboot open IMSM again, and carefully identify which of the two hard drives carries your data. This can only be done by identifying the correct SATA port. Remember when I said it was important to make notes at the beginning?
The first drive you fitted is the one carrying your data. Let's say it's connected to SATA4. The unused new drive let's say is connected to SATA3.
The final step is to migrate the data from SATA4 to SATA3 in my example.
Set this up in IMSM and a warning message will tell you the process can take a couple of hours. Ignore this if you're using 2TByte drives as it can take upwards of 8 hours.

Once the migration completion message has appeared you're done.
I used a pair of Seagate 2TByte hard drives and the capacity was given as 2,000,388,063,232 bytes or 1.81TB
The exact numbers will reflect how much your operating system bagged for itself and the precise amount of space the manufacturer sold you.


Return home