Solved

Spanned Volume - How disk writes are performed

Posted on 2013-12-13
10
2,116 Views
Last Modified: 2013-12-19
How would having a spanned volume help or hurt me in this scenario? I have 4 physicial disks comprised of 4 distinct volumes:
 
Disk 1, Volume 1 = system (250GB)
Disk 2, Volume 2 = data (2TB)
Disk 3, Volume 3 = backup (2TB)
Disk 4, Volume 4 = backup (2TB)

If I were to convert Disk 3 and Disk 4 to dynamic disks and span volume 3 across both, how will my backup data be written to that 4TB spanned volume? Does data get written to Disk 3 in the volume until it is full, then continue to Disk 4 in the same volume OR does data get written to either disk at random? My ultimate question or concern is that if I lose one physical disk in the spanned volume, what backup data will I be left with (if any)?
0
Comment
Question by:jcb431
  • 4
  • 3
  • 2
  • +1
10 Comments
 
LVL 33

Assisted Solution

by:PowerEdgeTech
PowerEdgeTech earned 250 total points
ID: 39718181
Although I think I once heard that it fills one, then "spills over" to the next, I wouldn't count on any data if either in the span gets lost, as it is not designed to be redundant at all:

http://technet.microsoft.com/en-us/library/dd163559.aspx
0
 
LVL 70

Assisted Solution

by:garycase
garycase earned 250 total points
ID: 39718390
According to Microsoft's documentation, "... , a spanned or striped volume will be broken and all data lost if any disk in the volume fails."

Only mirrored volumes and RAID-5 volumes (supported by Server OS's, but not Win7 or 8), with their built-in redundancy, support rebuilding after a failed physical disk.
0
 

Accepted Solution

by:
jcb431 earned 0 total points
ID: 39719478
I read up on the subject. You are both correct. Data is written sequentially (spills over) from disk to disk in a spanned volume as each disk approaches full capacity. It is not written in alternating fashion (interleaved) as is the case with striped volumes. A spanned volume will "break" if any disk within the volume fails. There is no replication or parity information stored in a spanned volume, therefore the volume cannot recover from even just one single disk failure. Now the answer to my question...

"If I lose one physical disk in the spanned volume, what backup data will I be left with (if any)?"

Essentially, all data that was written to the surviving disks will still exist on those surviving disks; however, because the volume is no longer existent, it will no longer mount in Windows as would be the case of a single disk failure in RAID 1. Instead, accessing that data will require the use of a third party disk utility to mount the surviving disks and recover the surviving data (i.e., http://www.quetek.com/). Windows does not natively provide this functionality. The data lost will be, not only any data on the inoperable hard disk, but also any data that overlaps a surviving disk and the dead disk that in effect constitutes usable information.

Being that this is the case, then why ever use a spanned volume? If you have JBOD you can create a spanned volume to take advantage of ALL the available storage capacity on the included disks, regardless of each individual disk's storage capacity. In contrast, a striped volume using JBOD is relegated to a total volume storage capacity equal to the SMALLEST individual disk's storage capacity.

Bottom line.... I ain't gonna do it. There is not a benefit in the scenario I painted. I'd have to have a third party utility to retrieve the data from the surviving disk of the spanned volume in the event of a disk failure AND the data retrieved would be incomplete. I'm marking this as the answer but giving you guys the points. Thanks!
0
Windows Server 2016: All you need to know

Learn about Hyper-V features that increase functionality and usability of Microsoft Windows Server 2016. Also, throughout this eBook, you’ll find some basic PowerShell examples that will help you leverage the scripts in your environments!

 
LVL 33

Expert Comment

by:PowerEdgeTech
ID: 39719537
"why ever use a spanned volume?"

I have never personally thought using this was a good option, so never had first-hand experience using or supporting it, but I once saw this on a Microsoft certification prep exam, where the answer was, in order to add space to a disk that was running low, to "span" to another disk.  I don't know how Microsoft figures this is a good feature, unless it was intended only to be an emergency/temporary fix, OR it is simply a lingering relic feature of a different time in computing when it might have been considered "adaptive" and "forward-thinking" for the time and options that were then-available.
0
 

Author Comment

by:jcb431
ID: 39719541
I'm in agreement. As per the hyperlink you provided... "Why, then, would anyone use spanning? Because they have hardware RAID to provide the redundancy. This combination offers the best of both worlds—redundancy provided by the hardware RAID controller and flexibility to expand volumes as needed, using Disk Management. Yet another compelling argument for hardware RAID, in case you needed any more."

The suggestion is that you can mirror a spanned volume, perhaps to a singular disk that has equal or greater storage capacity. Like you, I have no experience with this and can only suppose that the rationale for a spanned volume is what you suppose, an emergency/temp fix or a temporal solution.
0
 
LVL 70

Expert Comment

by:garycase
ID: 39719677
Why ever use it?   Simply to provide a larger volume than a single disk supports, but ONLY if you have good backups.

I never used them until recently (last year) when drive capacities became so large that I could store my media collection on a "single" volume in my main PC.    I created a 16GB volume with 4 4TB disks; and have a backup utility run every night that turns on my backup server; syncs the backup; and then shuts down the server.     Previously I ran the server 24/7 ... now it's on for just a few minutes/day.     I would NOT use spanned volumes if you do not have a complete backup, however.    With a full backup, a failed disk means you need to (a) replace the failed disk;  (b) recreate the spanned volume across all of the disks involved; and then (c) copy the entire backup to the spanned volume.   Clearly that's a LONG copy process if you have many TB's of data.
0
 

Author Comment

by:jcb431
ID: 39719696
Agreed. If you have a spanned volume, you need a backup of it. The scenario I described would mean using the spanned volume to backup the system and data volumes combined to a targeted spanned volume. In practice, using a spanned volume as a destination for backups is a piss-poor idea. Using them as a source for backup is a different story.

I enjoy this kind of discussion. I thank you both for being as engaging as you have been. I'm still a bit curious about using them in a RAID 1 scenario (as the source volume), but I've not found any documentation to support that notion. If either one of you do, please post.
0
 
LVL 70

Expert Comment

by:garycase
ID: 39719782
Windows does support replacing failed volumes if you're using dynamic disks with either mirrored (RAID-1) or RAID-5 volumes.    But only the server OS's support RAID-5.

It's a shame Windows 7 (and 8) don't support RAID-5 volumes with dynamic disks.   There's no technical reason they don't ... the capability is simply "crippled" to distinguish them from the server OS's.    [There are actually instructions you can find to enable this capability on the desktop OS's, but I do NOT recommend doing this -- you'd potentially be one Windows Update away from losing all your data !!]
0
 
LVL 55

Expert Comment

by:andyalder
ID: 39720062
Good that you have worked out why not to use spanned or striped disks in your situation but they were occasionally needed in the past when there was a 2TB limit on what Windows saw as disks. With large external RAID systems it was sometimes necessary to slice the array into several sub-2TB logical disks, present all of these slices to Windows and then stitch them together using dynamic disks. In these situations you could still lose a physical disk and keep going due to the underlying RAID on the external unit. Fortunately the 2TB barrier has now gone in most situations so such a bodge is no longer required.
0
 

Author Closing Comment

by:jcb431
ID: 39728789
I expanded significantly on the response of the two that helped me. I think that it is important for others seeking an answer to this question in the future to read the the full solution. I awarded all of the points to both of the respondents.
0

Featured Post

Backup Solution for AWS

Read about how CloudBerry Backup fully integrates your backups with Amazon S3 and Amazon Glacier to provide military-grade encryption and dramatically cut storage costs on any platform.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Concerto Cloud Services, a provider of fully managed private, public and hybrid cloud solutions, announced today it was named to the 20 Coolest Cloud Infrastructure Vendors Of The 2017 Cloud  (http://www.concertocloud.com/about/in-the-news/2017/02/0…
Many businesses neglect disaster recovery and treat it as an after-thought. I can tell you first hand that data will be lost, hard drives die, servers will be hacked, and careless (or malicious) employees can ruin your data.
This video Micro Tutorial explains how to clone a hard drive using a commercial software product for Windows systems called Casper from Future Systems Solutions (FSS). Cloning makes an exact, complete copy of one hard disk drive (HDD) onto another d…
This tutorial will walk an individual through the process of installing of Data Protection Manager on a server running Windows Server 2012 R2, including the prerequisites. Microsoft .Net 3.5 is required. To install this feature, go to Server Manager…

726 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question