Resizing Disk in Windows Server 2008 R2 SP1

[sidnuts]

Our server is a HP Proliant ML330 G6, with Windows Server 2008 R2 SP1 installed. We currently have 3x240gb harddisks installed onto a P410 Smart Array controller (RAID5).

 

Our problem is, we have just purchased 3x2tb harddisks with the hope of replacing the current 3x240gb disks. We have run a full backup job (using windows backup) of our server, replaced the 240gb disks with the 2tb disks and restored the backup. But when we check the size of the new disks it is showing only 240gb not the 4tb capacity we were expecting.

Is there a way we can resize the disks after restoring the backup?

[Dos_Box]

Are they RAID?

[sidnuts]

Thank you for the reply, yes on RAID5

[ChrisMiles]

If the problem is the partitions are simply the wrong size, you should be able to expand it using Disk Management.

 

If windows shows the actual disk (not drive/partition) as the 240gb then you will probably need to destroy the array and make a new one from within the raid controller bios and re-restore the backup.

[sidnuts]

Thanks for the reply Chris, posted this on experts exchange too, replies were pretty much the same as yours. Was advised to stay away from RAID5 though and go down the route on RAID10. Think we'll have a go at this over the summer hols. Got inspection in a couple of weeks so will leave things as there are until then.

[SYNACK]

Pfft, RAID10 is for people with to much money or not enough sense to buy more cache. RAID5 is not that much slower and is much more efficient with disk space. If you are going for extra redundancy go with RAID6. Faster access, buy more cache for the controller and/or faster/more drives.

[ChrisMiles]

Pfft, RAID10 is for people with to much money or not enough sense to buy more cache. RAID5 is not that much slower and is much more efficient with disk space. If you are going for extra redundancy go with RAID6. Faster access, buy more cache for the controller and/or faster/more drives.

 

Whilst that might be somewhat true for file storage servers in certain environments, with I/O intensive applications, especially those that pretty much exclusively require random disk access such as SQL and Exchange where cache really isn't going to help much, raid 5 is significantly slower and while I have seen exchange servers running on raid 5 before with no noticeable slowdown, this is usually in small networks where the server is normally overspeced anyway.

 

The most optimal setup for exchange imo is raid 10 for databases and either a raid 1 or raid 10 depending on your budget for logs. Since the logs are pretty much only accessed sequentially, high random access is not really necessary unless the server is usually under extremely high load and extra write performance is needed.

 

Raid-5 doesn't really suffer from any read performance issues, however, when it comes to disk writes, the parity calculations and what is effectively a minimum of 3 physical disk writes that is needed, it leaves a lot to be desired. And with the price of storage these days anyway, it really isn't worth sacrificing the performance to gain some extra storage, especially on an exchange server.

 

Anyway, that's just my opinion.