• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Windows Home Server and expandable storage

jtvang125

Diamond Member
From what I've read so far WHS will take all your drives and combine them into one large one. Does this suffer the same problem with spanning volumes where if one drive goes down the whole volume is lost? Also how big can this volume be? I'm assuming that depends if you format the volume using MBR or GPT right? Also how is the support for hardware raid?
 
WHS monitors your drives and is supposed to alert you if a drive is in danger of failure, I haven't run into this issue yet /*crosses fingers*/ but I know that it has this feature. What is supposed to happen if you do have a drive failure is that WHS will identify the drive and tell you to remove it from the storage pool, if you have enough space to remove the drive none of your data will be lost. If you are in a situation where there isn't enough free space to remove the drive I believe you will have to add more storage before you can remove the bad drive without risk of data loss. Hope that makes sense.

Not sure on the size limit, I have about 5TB right now using a mix of internal and external drives right now.
 
If a drive goes bad, you lose all the data on that drive. Any data on another drive will be fine, and WHS will remove the file tombstones for lost data.

GPT is not supported. You need to use MBR. This puts a practical cap at 2TB right now.

Hardware RAID is not officially supported, but you can still use it.
 
Originally posted by: jtvang125
From what I've read so far WHS will take all your drives and combine them into one large one. Does this suffer the same problem with spanning volumes where if one drive goes down the whole volume is lost? Also how big can this volume be? I'm assuming that depends if you format the volume using MBR or GPT right? Also how is the support for hardware raid?
Virge's summary is great. There are reports of WHS installations greater than 20 TB. Until WHS supports GPT, the largest individual disk that can be used is 2 TB. The next version of WHS (2010?) will probably have support for physical disks larger than 2 TB.

As noted, if a secondary data disk fails, you can remove it from WHS and pull the hardware disk. You'll lose any data on the disk that wasn't part of a redundant folder. You can attempt standard data recovery operations on the disk, since it's a Basic disk with a single NTFS-formatted partition. If the System disk fails (along with the included primary Data partition), there's a recovery system built into WHS to put WHS on the new System disk and to re-catalog the remaining files.

I've simulated the failure of System and secondary data disks in WHS and it seems to work as advertised.
 
Originally posted by: RebateMonger
Originally posted by: jtvang125
From what I've read so far WHS will take all your drives and combine them into one large one. Does this suffer the same problem with spanning volumes where if one drive goes down the whole volume is lost? Also how big can this volume be? I'm assuming that depends if you format the volume using MBR or GPT right? Also how is the support for hardware raid?
Virge's summary is great. There are reports of WHS installations greater than 20 TB. Until WHS supports GPT, the largest individual disk that can be used is 2 TB. The next version of WHS (2010?) will probably have support for physical disks larger than 2 TB.

As noted, if a secondary data disk fails, you can remove it from WHS and pull the hardware disk. You'll lose any data on the disk that wasn't part of a redundant folder. You can attempt standard data recovery operations on the disk, since it's a Basic disk with a single NTFS-formatted partition. If the System disk failes (along with the included primary Data partition), there's a recovery system built into WHS to put WHS on the new System disk and to re-catalog the remaining files.

I've simulated the failure of System and secondary data disks in WHS and it seems to work as advertised.

I'll second this. I had a non-simulated disk failure and recovery options worked perfectly. I only lost about 15gb of data from the 250gb drive that went bad, though.
 
The way WHS uses hard drives is quite elegant. Even though it combines them all into one volume, WHS will actively control what files reside on which drives, so you really get nice redundancy. The trick is to give WHS enough hard drives to do it efficiently. Two drive systems are adequate, but a 3 or 4 drive system will make WHS's job alot easier.

Its best you don't use any hardware raid on a WHS system, because WHS gets its strength from having access to multiple volumes in order to mirror files properly. If WHS can only see one volume, your basically cutting its legs out from under it and turning it into a glorified filing cabinet. Plus, who needs the hastle of spending tons of cash on raid gear when it doesn't even really help in this case. Your better off taking that money and spending it on 4 or more large hard drives.
 
Originally posted by: MStele
Its best you don't use any hardware raid on a WHS system, because WHS gets its strength from having access to multiple volumes in order to mirror files properly. If WHS can only see one volume, your basically cutting its legs out from under it and turning it into a glorified filing cabinet. Plus, who needs the hastle of spending tons of cash on raid gear when it doesn't even really help in this case. Your better off taking that money and spending it on 4 or more large hard drives.
I definitely agree about RAID underneath WHS. While I can see some logic behind a two-disk RAID 1 volume to host the WHS System disk, putting RAID beneath additional data disks reduces the disk flexibility of WHS and makes recovery more complicated.

WHS, by nature, is REALLY simple to use, manage, and expand. And it already has folder-level disk redundancy built in if you want it. Throwing in RAID causes the simplicity and ease of expansion to disappear.
 
Originally posted by: RebateMonger
Originally posted by: jtvang125
From what I've read so far WHS will take all your drives and combine them into one large one. Does this suffer the same problem with spanning volumes where if one drive goes down the whole volume is lost? Also how big can this volume be? I'm assuming that depends if you format the volume using MBR or GPT right? Also how is the support for hardware raid?
Virge's summary is great. There are reports of WHS installations greater than 20 TB. Until WHS supports GPT, the largest individual disk that can be used is 2 TB. The next version of WHS (2010?) will probably have support for physical disks larger than 2 TB.

As noted, if a secondary data disk fails, you can remove it from WHS and pull the hardware disk. You'll lose any data on the disk that wasn't part of a redundant folder. You can attempt standard data recovery operations on the disk, since it's a Basic disk with a single NTFS-formatted partition. If the System disk fails (along with the included primary Data partition), there's a recovery system built into WHS to put WHS on the new System disk and to re-catalog the remaining files.

I've simulated the failure of System and secondary data disks in WHS and it seems to work as advertised.

So basically if I don't have replication on for a folder and a drive goes down I'll just lose data on that drive correct? Also since WHS spreads out the data among the pool of drives how would I know which files actually are lost in the event of a drive failure since WHS doesn't show which physical disk it wrote the data to? One last question, the 2 TB limit is a physical limit right? Since I have 10 drives with about 9tb usuable I'll see one 9tb logical volume and not a bunch of 2tb ones?

Thanks for the help and sorry for all the questions?
 
Originally posted by: jtvang125
So basically if I don't have replication on for a folder and a drive goes down I'll just lose data on that drive correct? Also since WHS spreads out the data among the pool of drives how would I know which files actually are lost in the event of a drive failure since WHS doesn't show which physical disk it wrote the data to? One last question, the 2 TB limit is a physical limit right? Since I have 10 drives with about 9tb usuable I'll see one 9tb logical volume and not a bunch of 2tb ones?

Thanks for the help and sorry for all the questions?

Yes, you would only lose the (non-replicated) data on the failed drive. There is an add-on that will show you which files are on which drive. It would be sort of a hassle to use it for figuring out which files are lost, but maybe there's a better way I'm not aware of.

The 2TB limit is on the physical drives, yes, and only because WHS is based on Server 2003, and only supports MBR disks, and not GPT. Your share would appear as having 9TB usable space, in your example. You don't really get a 'logical volume' as you would normally think of in Windows, just a mass of space available via shares.
 
Originally posted by: ZetaEpyon
There is an add-on that will show you which files are on which drive. It would be sort of a hassle to use it for figuring out which files are lost, but maybe there's a better way I'm not aware of.
Determing WHICH files are lost is a problem. Just like in any other file server. I've talked to some WHS Microsoft people and there's no good way to do this.

After WHS is repaired, you can look at your shares and you'll have to figure out what files are gone. Because of how WHS tries to put "appropriate" files in the same place, a bunch of files added at the same time would, hopefully, be near each other. But there's no guarantee.

Unfortunately, Windows doesn't even have a simple way to make a directory-by-directory printout of files on a PC, Home Server or not.

As with any file server, the best way to "recover" data is to have recent backups and just restore all the files from backups.
 
If your properly setting up your backups for redundancy there should never an issue about which files are lost and which aren't, because you would have specific directorys setup for temporary storage and the rest for permanent (reduandant) storage. My personal opinion is to avoid putting misc files on the server at all unless your moving them from point to point, but thats me. Thats what extra harddrives on your workstation are for.
 
Back
Top